en con
TRANSCRIPT
Energy efficiency optimization for Power PlantsTraining Program – Dr.G.G.Rajan
1
ENERGY EFFICIENCY OPTIMIZATION FOR POWER
PLANTS 3 DAY TRAINING PROGRAM 19th to 21st December ‘11
Organised byNational Thermal Power Corporation
Vindhyachal
Course Director Dr.G.G.Rajan / Kochi
India
Energy efficiency optimization for Power PlantsTraining Program – Dr.G.G.Rajan
2
Program objective
Existing energy scenario all over the world warrants effective control on generation,
utilization and conversion at all levels. Energy utilization must be done more judiciously than the
past, without loss of product quality / quantity and environmental factors. This could be achieved
only by optimization, a special technique involving mathematics, statistics and Operations
research.
The program will explain the concepts of energy efficiency optimization with appropriate
examples. At the end of the program, participants will be able to appreciate the role of energy
efficiency optimization and it’s practical application.
Many examples covered in the program are real life cases and hence participants may
compare their live problems and adopt a suitable problem solving approach.
I thank the program organizers Messrs NTPC,Vindhyachal,India for their efforts in
organizing this program at short notice.
Dr..G.G.Rajan
Energy efficiency optimization for Power PlantsTraining Program – Dr.G.G.Rajan
3
CV OF Dr.G.G.Rajan
.
Dr.G.G.Rajan is a Chemical Engineer with qualifications in Mathematics, Statistics,
Management, Operations Research and Computer Applications. With his vast ‘hands-on’
experience and expertise in Refining, Petrochemical and Fertilizer industry, He had developed a
number of process models and had published and presented about 150 papers in national and
Inter National Seminars related to energy , environment , utility management, process
decisioneering etc. He has written a book titled ‘Optimizing Energy Efficiencies in Industry’ ,
Published by Tata McGraw Hill and McGraw Hill USA. He has also written a book titled
‘Practical Energy Efficiency optimization’ published by PennWell,US in 2006.
Rajan.G.G. the Chief Executive ,Techno Software International
Kochi, India since February 2002 was holding the position of
DGM (R&D) in Kochi Refineries Ltd.
Prior to this, he was the Chief Manager in Energy &
Environment section of KRL.He was deputed to Centre for
High Technology , Ministry of Petroleum, New Delhi, India as
an Additional Director in 1988, for six years to accomplish
certain special assignments
Energy efficiency optimization for Power PlantsTraining Program – Dr.G.G.Rajan
4
He had won a number of awards for his concepts on Productivity Management. He had
been the Chairman of an Energy Audit Team in 1990s, set up by the Ministry of Petroleum
and Natural Gas for energy auditing of three major refineries ( Chennai Petroleum
Corporation, Kochi Refineries Ltd, HPCL Vizag refinery).
Besides this he had conducted energy auditing of petrochemical plants like Asian
Peroxide Ltd, HOCL,KRL-Aromatics Plant, BPC Aromatics Plant , Fine papers Dubai etc,
Based on this he had developed an auditing software called Technical Audit.
His area of specialization is Energy and Environment Auditing, Performance Evaluation,
Bench Marking , Productivity Analysis and Profit Maximization. By virtue of his contribution to
Technology, he has been selected as a member of International Who is Who Society, Gibralter.
His Software Techno Therm , Technical Audit and SCIMOD are very popular and extensively
used by many industries.
Email: [email protected]
URL: http://business.vsnl.com/ggrtech
Energy efficiency optimization for Power PlantsTraining Program – Dr.G.G.Rajan
5
Program Contents
1. Energy efficiency and productivity • Fuel System
• Boiler / Steam System
• Turbine / Power system
• Thermal Efficiency / Energy transmission
• Energy Losses
2. Optimization Basics • Applications for Power Plant operation
• What to Optimize and how
• Cost / benefit Analysis – Evaluation
3. Power Plant Management • Boiler System
• Utilization of Power Management System
* Cogeneration
• Transmission System
• Waste heat recovery
• Total Power Plant
4. Performance Monitoring Techniques • Boilers
• Steam turbines
• Heat recovery steam generators
• Loss control
5. Energy Conservation through energy efficient technology & equipment • Emissivity coatings
Energy efficiency optimization for Power PlantsTraining Program – Dr.G.G.Rajan
6
• Enriched air
• High efficiency air heater / burners
• On-line cleaning
• Boilers tuning control
• Turbine monitoring system
• CW/ BFW / CEP Pumps
• Fans & Blowers
• Super Heater
Summing up / Discussion / Q A Session.
Conclusion
Energy efficiency optimization for Power PlantsTraining Program – Dr.G.G.Rajan
7
ENERGY EFFICIENCY AND PRODUCTIVITY
1
Energy efficiency optimization for Power PlantsTraining Program – Dr.G.G.Rajan
8
ENERGY EFFICIENCY AND PRODUCTIVITY
Introduction:
All industries consume four resources namely men , machine , materials and money.Of
late Energy is also treated as a resource, though it forms the subset of money resource. Financial
performance of any enterprise is ultimately determined by how effectively these resources are
utilized and how much profit has been generated for the resource inputs, of which energy has a
major role to play. This evaluation is extremely complex as financial performance is linked to
productivity of labor, machine ,material and money (4 M's of management - men, machines,
materials and money) .
These resources being inter-convertible to a limited extent , an optimum resource mix
should be evolved for a given economic scenario for achieving cost-effectiveness of the industry
under consideration . Each resource has an associated cost . An example is a highly automated
industry which has less than 30% of the man-power compared to a conventional industry of
the same magnitude. and production facility .This could be achieved only because of high
degree of automation. This strategy involves heavy capital input . Hence, an economic analysis
of cost of savings in man-power against additional capital investment has to be evaluated and
justified .
Material resource ( comprising hydro carbon feed stoch, fuel etc ) ,is another cost
centre that has an impact on the financial performance of the industry. It is normally felt that
cheaper hydrocarbon feed stock shall generate higher profits for a given set of production facility
comprising primary unit and other conversion processes.. This may not be always true in view of
the fact that the yield pattern of cheaper feed stock are invariably inferior and necessitate high
degree of secondary processing ,treating and waste disposal .This tends to increase the operating
cost in terms of fuel and loss, pollution control measures, additional treatment cost etc.
Equipment performance (machine productivity) is an important factor that determines
the profitability of the industry. Organizational procurement procedures and policies necessitate
Energy efficiency optimization for Power PlantsTraining Program – Dr.G.G.Rajan
9
the choice of low cost equipment during the selection stage for new / existing processes. A
scientific method need to be applied in this area as cheaper equipment normally have lower life
cycle , reliability energy efficiency and high failure rates leading to enhanced operation,
maintenance and replacement costs which in turn affect the run length of the operating units.
This reflects on production loss, wasteful expenditure on maintenance and lower profitability.
This factor that has to be evaluated to arrive at right equipment maintenance / replacement
decision. This is a crucial managerial decision as the productivity of the enterprise is linked to
equipment availability, when other parameters remain constant
Capital performance (capital productivity ) is the ultimate objective of any enterprise
. This is normally determined by the production volume ( in monetary terms) to capital input.
For example when the performance of two industries are compared, quantity of products
produced per unit of capital input is considered as a measure of performance. For achieving
effective financial performance, one has to identify all the cost and profit centres , cost-intensive
elements and develop performance models for monitoring and control.
This document covers the most modern concept of Enterprise Resource Planning using some
Linear Programming and Scientific Models which are used by modern industries for
accomplishing effective cost control and high profitability within reasonable limits of
accuracy .
Energy efficiency optimization for Power PlantsTraining Program – Dr.G.G.Rajan
10
A. Productivity and Resources Utilization.
Productivity is the efficiency with which resources are used to produce goods and
services. This is generally expressed as the ratio of output to input. Most common productivity
ratios in use are labor productivity, Capital productivity, Material productivity and Machine
productivity.
Enterprise performance is the sum total of all productivity which reflects on how
effectively labor , capital, equipment and materials are deployed to achieve the output.
Hydrocarbon processing industry operations are very capital and energy intensive.
Consequently, partial productivity ratios pertaining to the efficiency of capital, energy and
feedstock play an important role in plant performance and cost effectiveness. Fig 1 shows the
concept of total and partial productivity which is used in conventional Enterprise Resources
Planning .
Energy efficiency optimization for Power PlantsTraining Program – Dr.G.G.Rajan
11
B. Energy Efficiency and productivity.
Since we are more concerned with Energy productivity, energy efficiency must be high to
achieve this objective and it may be represented by a simple relationship
Energy productivity may be increased by
1. Increasing the output for the same energy input
2. By reducing the energy input for the same output or
3. By increasing the output and simultaneously reducing energy input.
Though it looks simple in the above mathematical form, this is a complex problem in the
Hydro Carbon Processing Industry as the energy input to the process are in various forms such
as fuel, steam, power, thermal energy etc.
Energy intensity (EI) is a commonly used index to estimate energy efficiency (EE) for
countries, but it neglects the specific structure of energy consumption. For micro level analysis
and energy efficiency optimization, the problem may be broken down into smaller problems and
the concept of energy supply chain must be applied.
Fig 2 explains the concept of Energy supply chain. This refers to the case of a Thermal
power plant, which uses coal , fuel oil and fuel gas as the primary fuel to generate high pressure
steam. The super heated steam is supplied to the steam turbine which generates electricity.
Electric power is then transmitted by a transformer to the consumer’s area, located at a far off
place. This is shown in the fig2. Overall efficiency of the system is the product of individual
efficiencies of elements constituting them.
Fig2. Energy supply chain – Power Generation
Ep = Output
Energy input.
Power Boiler Consumer
TransformerStation
Turbo Generator
SteamTurbine
Energy efficiency optimization for Power PlantsTraining Program – Dr.G.G.Rajan
12
C. Energy efficiency vs Productivity - Fuel System.
Fuel system has an important role in maintaining energy efficiency of boilers. Different
fuels behave differently during combustion process. The most generic term used to identify
efficient fuel utilization is the fuel efficiency.
Fuel efficiency, in its basic sense, is the same as thermal efficiency, meaning the
efficiency of a process that converts chemical potential energy contained in a carrier fuel into
kinetic energy or work. Overall fuel efficiency may vary per device, which in turn may vary per
application, and this spectrum of variance is often illustrated as a continuous energy profile.
Non-transportation applications, such as industry, benefit from increased fuel efficiency,
especially fossil fuel power plants or industries dealing with combustion, such as ammonia
production during the Haber process.
Parameters affecting the fuel efficiency in the case of Process Heaters and Boilers are
Fuel type ( coal / liquid / Gaseous )
Fuel temperature
Fuel viscosity / particle size ( for solid fuels )
Air to fuel ratio ( for combustion )
Steam to Fuel ratio ( for atomization )
Burner condition etc
Fuel temperature and viscosity play an important role in combustion. Since viscosity and
temperature are interrelated, the viscosity of fuel at the burner tip is specified by the burner
manufacturers.
The production of energy by the fired heater consists of the burning of fuels that are
primarily of fossil origin. However, the variety of fuels is great when compared with other
industries. The liquid fuels burned ranges from liquid butanes and pentanes, to light fuel oils
such as diesel and #2, to heavy fuel oils such as #6, and many heavier liquids.
Gaseous fuels may come in an even wider variety. Fuels such as hydrogen, methane,
ethane, propane, butane and carbon monoxide are burned either separately or in an infinite
variety of blends which sometimes contain inert components such as nitrogen and carbon
dioxide. The heating value and specific gravity of the blends are quite variable. Lower heater
Energy efficiency optimization for Power PlantsTraining Program – Dr.G.G.Rajan
13
values ranging from lower than 100 Btu per cubic foot, to in excess of 3,000 Btu per cubic foot
are not unusual.
Further complicating the firing of a process heater are requirements such as flame shape,
radiant flux rates and unique sources of oxygen for the combustion reaction. Heat transfer for
different processes and different heater designs to accomplish the heat transfer for a particular
process have engendered the design of different burners that produce a multitude of flame
shapes. The flame shapes must be compatible with both the mechanical configuration of the
radiant combustion zone of the heater, and the radiant heat flux rate required by both the process
and heater design
Burner types consist of raw gas burners that mix fuel and air in the combustion zone, and
pre-mix burners that mix the fuel and air prior to the combustion zone for gaseous fuel firing.
There is a multitude of oil burners which use many methods of atomization for the fuel oil. In
addition, there are various combinations of the raw gas, pre-mix and liquid burners.
The burning of the various fuels described produce nitrogen oxides. The production of
nitrogen oxides has become a matter of concern to industry. First, NOx has an adverse effect on
the environment. Second, there is a growing volume of regulation dealing with the emission of
NOx to the atmosphere.
Process heaters are also somewhat unique in that the oxygen for combustion can come
from a variety of sources. In the majority of cases, the source of oxygen is atmospheric air
containing 21% oxygen by volume.
However, atmospheric air can be supplied in many forms. For many years, the most
frequent method of air supply to the burner was by natural draft with the air at ambient
temperature. More recently, the combustion air has been supplied at a positive pressure to permit
better control of the flame pattern and excess air.
Further requirements to conserve energy have brought about the use of preheated
combustion air. Because of the pressure to conserve fuel, other high temperature sources of
oxygen for the combustion reaction are now being utilized. For example, the off-gas from a gas
turbine can be used. This stream contains 15% to 16% by volume of oxygen at a temperature
usually in excess of 900° F. When a stream of this type is used as a source for oxygen, the
sensible heat of the TEG stream can be recovered. Another source of oxygen is the off gases
from certain kilns.
Energy efficiency optimization for Power PlantsTraining Program – Dr.G.G.Rajan
14
Steam and air assisted atomizers are widely used in oil burners in industrial furnaces or
boilers. In industrial burners, the fuel oil is delivered to the burner process by an atomizer
nozzle, also known as an “oil gun”, when referring to the fuel supply lance and oil atomizer
assembly as a whole. The quality of atomization will influence combustion performance because
combustion begins immediately downstream of the atomizer nozzle.
The combustion performance attributes that furnace operators are most interested in are
flame length, consumption of atomizing medium, turndown ratio, NOx emissions, and
particulate emissions. In the past, due to lack of a good way of measuring atomization
effectiveness, mostly empirical trial and error was used in industrial research to select the best
atomizer nozzle. Consequently the empirical approach was limited to the performance of
the atomizer nozzles that were readily available, and could not provide a method to design the
atomization specifically to achieve a given combustion performance target.
The study and comparison present findings about how the atomization parameters can
be varied to contribute improvement in specific areas of combustion performance.
A rule of thumb to determine the burner efficiency / performance is to compare the
combustion zone temperature of heater / boiler at different fuel temperatures , atomizing steam to
fuel ratio , air to fuel ratio and air temperature. This is a practical method to determine the
performance of the fuel system.
D. Case study :
Hyundai Oilbank Co., Ltd. needed to increase the processing capacity of its No. 2
Crude Distillation Unit located in Daesan, South Korea. When the operators increased the burner
heat release in the heater, by increasing oil flow to the burner, the flame length also increased
and created flame impingement.
Burner modernization in the heater by replacement of its conventional EA-type oil guns
with the new HERO oil guns having phased atomization of fuel oil to efficiently atomize the
liquid fuel with less steam consumption resulted in shorter flame lengths, reduced soot
formation, increased turndown and lower NOx emissions.
RESULTS
• Up to 15 % shorter flame lengths at design rate.
Energy efficiency optimization for Power PlantsTraining Program – Dr.G.G.Rajan
15
• 36 % less atomizing steam consumption; steam-to-oil ratio reduced from 0.40 to 0.27 kg-
steam/kg-oil.
• 5 % less NOx emissions; 210 ppm NOx before change and 200 ppm NOx after change.
• Hyundai Oilbank Co. will recover its capital investment of the 64 HERO guns over 15 months
in reduced energy costs alone.
Fig 3. Burner retrofitting results in energy savings
For a fired duty of 100 million kcal/hr and the heater efficiency of 87%, the fuel
consumption shall be 11495 kg oil /hr (10000 kcal/kg LCV). The atomizing steam consumption
between the two cases will be as shown in the table below.( Ref John Zinc )
Fired duty of heater in mmkcal/h
Heaterefficiency% on LHV
Fuelconsumption in kg/hr
Steam consumption kg/hr
Annualconsumption mt/yr
100 87 11495 4598 36784
100 87 11495 3104 24832
Savings due to modernization
1494 11952
This example shows the impact of burner efficiency on the heater performance in terms
of savings in atomizing steam and fuel consumption.
Energy efficiency optimization for Power PlantsTraining Program – Dr.G.G.Rajan
16
E. IMPACT OF BURNER AIR TO FUEL RATIOS
Periodic checking and resetting of air-fuel ratios is one of the simplest ways to get
maximum efficiency out of fuel-fired process heating equipment such as furnaces, ovens,
heaters, and boilers. Most high temperature direct-fired furnaces, radiant tubes, and boilers
operate with about 10 to 20 percent excess combustion air at high fire to prevent the formation
of dangerous carbon monoxide and soot deposits on heat transfer surfaces and inside radiant
tubes.
For the fuels most commonly used by the industry, including natural gas, propane, and fuel oils,
approximately one cubic foot of air is required to release about 100 British thermal units in
complete combustion. Exact amount of air required for complete combustion of commonly used
fuels can be obtained from the information given in one of the references.
Process heating efficiency is reduced considerably if the air supply is significantly higher
or lower than the theoretically required air.
Air-gas ratios can be determined by flow metering or flue gas analysis. Sometimes, a
combination of the two , works best. Fig 4 gives fuel lost as dry flue gas as % of input fuel for
various stack temperatures., which may be controlled by adjusting air-fuel ratios.
The excess air curves are labeled with corresponding oxygen percentages in flue gases.
To figure potential savings, you need to know:
• The temperature of the products of combustion as they leave the furnace
• The percentage of excess air or oxygen in flue gases, at which the furnace now operates
• The percentage of excess air or oxygen in flue gases, at which the furnace could operate.
On the chart, determine the available heat under present and desired conditions by reading up
from the flue gas temperature to the curve representing the excess air or O2 level; then, read left
to the percentage available heat (AH).
Calculate the potential fuel savings:
% Fuel Savings = 100 X ((%AH Desired - %AH Actual ) / %AH Desired)
Energy efficiency optimization for Power PlantsTraining Program – Dr.G.G.Rajan
17
F. FACTORS AFFECTING EXCESS AIR LEVEL REQUIREMENTS
Combustion systems operate with different amounts of excess air between high and low fire.
Measurement of oxygen and combustibles such as carbon monoxide in flue gases can be used to
monitor changes in excess air levels.
For most systems, 2 to 3 percent oxygen with a small amount of combustibles—only 10
to 50 parts per million—indicate ideal operating conditions. Processes that evaporate moisture or
solvents need large amounts of excess air to dilute flammable solvents to noncombustible levels,
to ensure adequate drying rates, and to carry vapors out of the oven.
Lowering excess air to minimal levels can slow down the process and create an
explosion hazard.
Fig 4. % Oxygen in Stack vs Heat Loss % in dry flue gas
Energy efficiency optimization for Power PlantsTraining Program – Dr.G.G.Rajan
18
G. BURNER SPECIFICATIONS
These burner specifications can improve fired heaters performance. Specifying the right
requirements for fired heater burners can improve the heater operation and reduce maintenance.
With the right type of burners, fired heater capacity can be increased by 5 to 10% and
thermal efficiency by 2 to 3 %.
Burner selection and specification should be done carefully as they have a direct impact
on heater operation. Table given below lists the major parameters for a good burner design.
TABLE —Burner selection criteria:
Burner type.
Users should specify only burner types that they have sufficient operating experience
with. Alternatively, the burner that has been proven in the industry elsewhere, or at least has
been successfully tested at the vendor's furnace test rigs under simulated conditions, should be
specified. A number of cases have been reported of severe production losses and shut downs due
to selection of unproven burners. It may appear to be costly to ask for testing of burners, as one
of the requirements, but it will prove much cheaper in the long run. This is particularly important
Burners should provide stable combustion with:
1. Ability to handle wide range of fuels
2. Provision for safe ignition
3.Easy maintenance
4.Good turn down ratio
5.Well defined flame pattern with all fuels and firing rates
6.Low excess air operation
7. Low noise and NOx levels.
Energy efficiency optimization for Power PlantsTraining Program – Dr.G.G.Rajan
19
now with new low Nox burner designs that may require higher excess air and have longer flame
lengths than normal burners.
Heat release and turndown.
The burner supplier should be able to meet varying demands of the user arising from his
process design and operation philosophy. Burners are normally designed to provide 120% of
their normal heat liberation at peak heat duty.
The user should specify very clearly the normal, maximum and minimum heat release
requirements of the burner to the vendor.
Over sizing burners normally leads to over firing of the heaters. Turndown capability of
these burners depends upon the type of oil gun selected and available oil and steam pressures.
The lower limit depends upon the ability to keep a stable flame with minimum oil throughout.
Air supply.
The availability of furnace draft determines the size of air registers. In case of a natural
draft burner, draft available at the burner must be calculated accurately from the firebox
dimensions and flue gas temperature, and specified clearly. If the draft specified is lower than the
available draft, it will result in oversized burners.
On the other hand, if higher draft is specified than what is actually available, burners will
not be able to give maximum heat release. If preheated air is used for combustion then
combustion air temperature at the burner needs to be specified to apply temperature correction
for the pressure drop. If forced draft burners are also required to operate under natural draft
conditions it must he dearly specified.
The natural draft conditions are controlling in such cases. Burner air supply should be
protected against variations in wind velocity, which may cause blowback. Forced draft burners
should be provided with adjustable dampers and a wind box gauge to ensure equal distribution of
air. Uniform air distribution to the burner can also be ensured by proper duct design or flow
modeling.
Energy efficiency optimization for Power PlantsTraining Program – Dr.G.G.Rajan
20
The burner wind box and front plate should be made out of at least 3-mm CS plates. This
is to prevent warping of plates and improve long-term operability. Roller bearings should be
provided for easy register movement and a positive locking device should be included to prevent
vibration from changing register position. Register air controls should be easily accessible
Excess air.
For complete combustion, it is necessary to supply enough excess air. A good burner
design calls for minimum possible excess air compatible with process requirements.
TABLE 2-Excess air levels
Normal excess air levels for different fuels are given in Table 2. As a rule of thumb every
10% extra air used in combustion translates into a loss of 0.7% in terms of efficiency. At
turndown conditions, which are mostly overlooked, the excess air requirement goes up
substantially
Fuel specifications.
The burner design is linked directly with the fuels to be fired. Some of the properties that
need to be listed very clearly for liquid fuels are fuel specific gravity, lower heating value, fuel
oil viscosity, fuel pressure, and temperature available at burner.
For gaseous fuels, molecular weight is also important. Fuel oil viscosity and pressure
available will govern the atomizer and oil tip design. If the fuel contains abrasive particles
hardened, oil tips should be specified.
Burner Type Fuel oil Fuel gas
Natural draft 15 to 25 10 to 20
Forced draft 10 to 15 5 to 10
Forced draft (preheated air) 5 to 10 5
High intensity 5 5
Energy efficiency optimization for Power PlantsTraining Program – Dr.G.G.Rajan
21
COAL FIRED BOILERS
In case of coal fired boilers, quality of coal especially the ash and sulfur content plays
an important role in maintaining boiler efficiency. Following figure shows the impact of ash
content on boiler efficiency.
ASH CONTENT IN COAL vs BOILER EFFICIENCY (LHV)
Besides ash content, other parameters which affect the coal fired boiler efficiency are the
particle size of coal. Pulverization is an energy intensive process . Hence the impact of these
parameters will have to be evaluated in coal fired boilers in the optimization study.
82
83
84
85
86
87
88
'16 % ' '21 %' '30 %' '35 %'
% eff blr
Energy efficiency optimization for Power PlantsTraining Program – Dr.G.G.Rajan
22
H. BOILER / STEAM SYSTEM.
Steam is consumed in process and power industries for driving compressors, turbines and
for heating process fluids ,creating vacuum by ejectors etc. Steam pressure depends on the
process configuration, type of equipments used and the choice of Combined Heat Power cycle
selected. In normal practice, cost of steam alone constitutes about 60 to 70 % of the total
utility cost.
Steam Consumption could be monitored by an overall steam balance of the set up on a
day to day basis which include the quantity of steam generated and steam consumed at various
sections. The difference between the two denotes steam loss which should be minimum.
Steam consumption is dynamic in nature and it's demand is a function of load factor of
individual unit , efficiency of various rotating equipments deployed, heat recovery level , feed
stock quality, operating severity etc of each unit / section .
Under these dynamic conditions, computer-based models could be used effectively to
monitor the steam consumption of individual unit, monitor overall steam demand and identify
deviations and problem areas for remedial action. A typical utility flow diagram of a process
unit is given in fig 5. In case of boiler system, boiler efficiency determines the quantity of steam
generated per ton of fuel burnt and the productivity parameter is the steam by fuel ratio. All
parameters remaining constant, Steam / Fuel ratio or specific fuel consumption is the key
performance indicator.
Applying Energy supply chain concept, steam generation quantity will depend on steam
and power demand of the total process. When process efficiency , power generation efficiency
and steam distribution efficiency are lower than the target, the specific consumption of fuel will
tend to go up.
Example :
A boiler generates 100 t/hr of steam at 65 kg/cm2g pressure and generates 10 mw power
in a back pressure turbine. Exhaust steam is 100 t/hr at 15 kg/cm2g pressure. Entire BP steam is
consumed in the process when the line loss is about 2.5%. After years of operation, line losses
increased to 4.0 % and process consumption increased by 5%. Under this condition steam
Energy efficiency optimization for Power PlantsTraining Program – Dr.G.G.Rajan
23
generation also increases to 106.6. t/hr, assuming no change in turbine efficiency. In this case the
productivity has come down by 6.6%.
Fig 5. Typical Steam , Power and Compressed Air System
If the turbine efficiency has also deteriorated during the period, the steam demand will
increase further. Assuming the steam / fuel ratio of the base case as 16.5 and the current ratio at
15.4, the increase in fuel consumption will be 106.6/15.4 – 100/16.5 = 0.8641 t/hr.
Typical energy supply chain for rhe boiler and steam system may be as shown below.
Energy efficiency optimization for Power PlantsTraining Program – Dr.G.G.Rajan
24
Boiler efficiency is a function of fuel type, atomization efficiency, air / fuel ratio, air
temperature, draft profile, convection and radiation losses, boiler shell condition, water tube
condition, etc.
In case of coal fired boilers, efficiency will be determined by pulverized coal size, ash
content, coal burner efficiency etc.
Similarly for the steam system, the consumption is a function of transmission efficiency,
temperature / pressure, equipment efficiency , line losses etc.
All these parameters will have to be checked thoroughly when either the boiler or steam
system efficiency starts dropping down.
I. TURBINE AND POWER SYSTEM.
Electric Power is generated by a turbo generator in which the heat energy in steam is
converted into kinetic energy, which rotates the rotor of the Generator, which in turn generates
power. In the case of gas turbine, fuel is burnt in a combustor and the hot gases expand in the
turbine section to convert thermal energy into kinetic energy which drives the rotor of the
generator. Power generated at the terminal is linked to the efficiency of the turbine and the
generator.
Efficiency of the rotating machine is linked to
Existing machine life or years of operation
Wear and tear of various components
Condition of stator / rotor coils and
Condition of bearings
Alignment etc.
Starting from the consumer end, the energy demand at turbine terminal may be evaluated
as shown in fig 6.
Energy efficiency optimization for Power PlantsTraining Program – Dr.G.G.Rajan
25
Following table shows the impact of energy efficiency of individual energy link on the
overall energy efficiency of the turbine and power system.
Consumer side demand 1000 kw
Above table shows the deterioration in overall energy efficiency of the system over a
period of operation or life cycle.
J. THERMAL EFFICIENCY AND ENERGY LOSSES
In any fuel fired equipment like heater / boiler / gas turbine etc, thermal efficiency refers
to the ratio of energy consumed for heating the process fluid or quantity of heat absorbed for
steam generation against the thermal energy input in the form of fuel or any thermal energy. This
may be represented as given below for a system
Thermal Efficiency may be mathematically expressed as
Case Turbine efficiency
Generatorefficiency
Transmissionefficiency
Consumerequipmentefficiency
Overallenergy
efficiencyBase 60.0 95.0 97.5 65.0 36.12 After 5 yrs 58.0 92.5 96.0 61.0 31.41 After 10 yrs 56.0 91.0 95.0 58.0 28.07 After 15 yrs 54.5 89.5 93.8 56.5 25.85
SYSTEM Energyinput
Energyutilized
EnergyLosses
Energy efficiency optimization for Power PlantsTraining Program – Dr.G.G.Rajan
26
Thermal = Energy utilized for doing work / Energy input (1)
= ( Energy input – energy losses ) / Energy input (2)
Thermal may be increased by reducing the energy input for the same energy
utilization / output or increasing useful energy output for the same energy input. This needs a
total system analysis and an energy loss break-up.
In the case of a boiler for example, energy input ( energy supply chain ) must be analyzed
scientifically as shown by the energy supply chain diagram. From the energy system link, it may
be noted that the useful energy input is determined by fuel quality, fuel atomization, excess air,
moisture content etc.
IMPACT OF FUEL QUALITY / MIX
In Thermal Power plants, the conventional fuel used is coal. Coals used as fuel in boiler
plants have high ash content. Impact of ash content on Boiler efficiency is as shown below.
ASH CONTENT IN COAL vs EFFICIENCY (LHV)
82
83
84
85
86
87
88
'16 % ' '21 %' '30 %' '35 %'
% eff blr
Energy efficiency optimization for Power PlantsTraining Program – Dr.G.G.Rajan
27
PARAMETERS USED IN EFFICIENCY CALCULATION
Fuel mix options for coal fired boilers
o Selecting low ash coal
o Blending Anthracite coal with high ash coal before crushing and pulverizing
o Blending raw petroleum coke with high ash coal and pulverizing the mix
Impact of fuel atomization / coal pulverization:
Fuel atomization / coal pulverization has direct impact on thermal efficiency of fired
heaters / boilers. Unburnt hydro carbons will be eliminated, when the degree of atomization is
high. With good atomization, air/fuel ratio requirement will be lower and the combustion zone
temperature also will be high. Hence there will be a reduction in fuel consumption and increase
in efficiency due to low heat loss from excess air.
Impact of excess air :
o As boiler excess air increases, efficiency of fired equipment like heater / boiler reduces.
o This is equipment specific and varies with the operating parameters such as fuel
composition, % Oxygen in flue gas and stack temperature.
o Figure given below shows the relationship for a specific case.
Component Wt % Wt % Wt % Wt %
Carbon 59.0 54.0 45.0 40.0
Hydrogen 3.0 3.0 3.0 3.0
Moisture 00.0 00.0 00.0 00.0
Oxygen 19.4 19.4 19.4 19.4
Sulfur 0.4 0.4 0.4 0.4
Nitrogen 2.2 2.2 2.2 2.2
Ash 16.0 21.0 30.0 35.0
Energy efficiency optimization for Power PlantsTraining Program – Dr.G.G.Rajan
28
Excess air vs Efficiency
Typical energy losses from a fired heater / boiler.
a.Dry Gas Loss : 7.3617
b.Air Moisture loss : 0.0000
c.Combustion Moisture Loss : 7.5381
d.Fuel Moisture Loss : 0.7614
e.Radiation Loss : 1.2100
f.Blow Down Loss : 0.5100
g.Unaccounted Loss : 0.2000
g.Loss due to combustibles : 0.0000
Total Loss % ( dry basis) 17.5813
k. ENERGY TRANSMISSION :
In any fired equipment like a heater / boiler etc, there exists an Energy transmission chain
from the point of energy system input to the consuming end. A typical energy supply chain for a
boiler system is as shown below.
85
86
87
88
89
90
30% 25% 20%
Boiler Eff%
Energy efficiency optimization for Power PlantsTraining Program – Dr.G.G.Rajan
29
Typical Energy transmission chain
Heat of combustion between fuel and air, generates hot gases in the furnace section of
the boiler. Hot gases will be at a temperature between 900 to 1000 oC.
Energy transfer takes place between the hot gases and the circulating water in the water
wall tubes. This energy transfer produces saturated steam in the boiler drum.
The hot gases after delivering energy to the BFW in the boiler leaves at a temperature of
450 to 500 oC in the flue gas duct and enters the steam super heater as shown below.
Fuel
Air
Boiler
Air pre heater
SteamSuper heater
Economizer
Air
Flue gas to stack
Hot air
BFW
Energy efficiency optimization for Power PlantsTraining Program – Dr.G.G.Rajan
30
Flue gas temperature is now lowered to around 350 oC. Energy available in the flue gas
is utilized to pre heat the air for combustion as shown in the figure below.
Air enters the shell side of the APH while flue gas passes through tubes as shown in
the figure. This increases thermal efficiency of the boiler by 9 to 11 %
Typical Air pre heater Layout.
Energy transmission in Economizer:
Energy efficiency optimization for Power PlantsTraining Program – Dr.G.G.Rajan
31
Main objective of economizer is to increase the thermal efficiency of the boiler by pre
heating boiler feed water by exhaust flue gases as shown in the figure
From the above figures, it may be noted that energy efficiency of the total system
increases by efficient energy recovery in various sections, so that energy loss is minimized. In
the case of any fired equipment, energy losses occur in the form of exhaust flue gases, as shown
in page 28. Fig given below shows the energy loss analysis of a typical boiler.
0.00 1.00 2.00 3.00 4.00 5.00 6.00 7.00 8.00
a.Dry Gas Loss :
b.Air Moisture loss :
c.Combustion Moisture Loss :
d.Fuel Moisture Loss :
e.Radiation Loss :
f.Blow Down Loss :
g.Unaccounted Loss :
g.Loss due to combustibles :
energy loss%
energy loss%
Energy efficiency optimization for Power PlantsTraining Program – Dr.G.G.Rajan
32
Exercise :
A power plant has 4 nos of coal fired boilers and 4 steam turbines to generate power.
Capacity of each boiler is is 750 t/hr and the capacity of each turbine is 120 mw with one
standby. One boiler and one turbine are standby. Coal consumption of each boiler is 135 t/hr.
Power demand is 300 mw. How will you operate the system such that the fuel consumption is
minimum ?
Energy efficiency optimization for Power PlantsTraining Program – Dr.G.G.Rajan
33
Exercise :
Following table shows the efficiency data of a steam turbo generator. What are your
observations and what are your suggestions for improvement ?
Steam turbine operating data
Month HP steam t/hr
Steam temp Power gen in mw
1 750 420 100 3 750 415 98.5 5 800 415 112.8 7 800 420 120.0 9 780 420 103.0 11 780 415 101.5
Energy efficiency optimization for Power PlantsTraining Program – Dr.G.G.Rajan
34
Notes :
35 Energy efficiency Optimization for Power plants. Faculty : Dr.G.G.Rajan
OPTIMIZATIONBASICS
2
36 Energy efficiency Optimization for Power plants. Faculty : Dr.G.G.Rajan
2. OPTIMIZATION BASICS
Optimization refers to the scientific management of operating / operational
parameters, which are within the control of the organization, to maximize productivity,
profitability and performance of the organization, within the imposed internal / external
constraints. In any business activity, a number of conflicting parameters exert their
influence on the product / services provided by the organization. Hence the organization
chief has to take appropriate decision on what has to be done to maximize the
performance at a given situation. This leads to an optimization problem(s), which are
made up of three basic ingredients:
An objective function which one wants to minimize or maximize. For instance,
in a power plant where a number of boilers are operating in open cycle and co generation
mode, the objective function is to generate power to meet the demand at lowest operating
cost and maximum profit. By using experimental data, it is possible to develop a user-
defined model, which will minimize the total deviation of observed data from predicted
model output. In this problem, the variables include the boiler capacity, turn down ratio,
turbine load, turbine efficiency, fuel cost, steam cost, total operating cost at a particular
mode of operation etc. A set of constraints that allow the unknowns to take on certain
values but exclude others. In this power plant problem, the constraints are the individual
boiler loads, it’s maximum efficiency point, turbine load, it’s efficiency point etc .
The optimization problem is then narrowed down to find the values of the
variables that minimize or maximize the objective function as required, while fulfilling
all the constraints.
A. Objective Function
Almost all optimization problems have a single objective function. The two
interesting exceptions are:
37 Energy efficiency Optimization for Power plants. Faculty : Dr.G.G.Rajan
B. No objective function.
In some cases (for example, design of integrated circuit layouts), the goal is to
find a set of variables that satisfies the constraints of the model. The user does not
particularly want to optimize anything so there is no reason to define an objective
function. This type of problems is usually called a feasibility problem.
Multiple objective functions.
Often, the user would actually like to optimize a number of different objectives at
once. For instance, in the power plant problem, the objective function is to minimize
operating cost and maximize power generation quantity simultaneously.
Usually, the different objectives are not compatible; the variables that optimize
one objective may be far from optimal for the others. In practice, problems with multiple
objectives are reformulated as single-objective problems by either forming a weighted
combination of the different objectives or else replacing some of the objectives by
constraints. These approaches and others are described in our section on multi-objective
optimization.
Variables
These are essential. If there are no variables, we cannot define the objective
function and the problem constraints.
Constraints
If Constraints are not present, this results in the field of unconstrained
optimization. In real life situations, all problems really do have constraints.
Example 1 Three boilers B1,B2 & B3 are in operation in a process unit, generating 100, 150
and 180 t/hr of steam respectively. Steam / Fuel ratio which is function of capacity
utilization of each boiler is given below. The design capacity of the boiler B1,B2 & B3
are 120, 180 & 210 tons/hr respectively. Determine the optimal load, the boilers should
handle to meet the steam demand of 430 t/hr at minimum fuel consumption. Fig 2.01
38 Energy efficiency Optimization for Power plants. Faculty : Dr.G.G.Rajan
shows this lay out. The flow diagram shows the normal operation which was based on the
decision of the plant superintendent.
The above data will be converted into fuel / steam ratio for each boiler and a
model developed between fuel consumption and steam produced. Table 2.01 shows the
above information in terms of steam produced in tons/hr versus fuel in kg. These values
will be used in boiler loading optimization model.
Table 2.01. Steam Generation Vs Fuel consumed
Boiler 1 Boiler2 Boiler 3
steam gen fuel steam gen fuel steam gen fuel
72 6000 108 8244 126 9130
84 6720 126 9333 147 10208
96 7385 144 10286 168 11200
108 8000 162 11408 189 12194
120 8571 180 12329 210 13125
Boiler 60% 70% 80 % 90% 100%
Boiler 1 12 12.5 13.0 13.5 14.0
Boiler 2 13.1 13.5 14.0 14.2 14.6
Boiler 3 13.8 14.4 15.0 15.5 16.0
Steam / fuel ratio of boilers at various loads.
BOILER 1 Capacity 120 t/h
BOILER 2 Capacity 180 t/h
BOILER 3 Capacity 210 t/h
Fig 2.01. Boiler Load optimization problem
430 t/hr demand
100 t/hr 150 t/hr 180 t/hr
39 Energy efficiency Optimization for Power plants. Faculty : Dr.G.G.Rajan
Note: Steam generation in tons/hr . Fuel Consumption in kg/hr
From the above data regression equations have been developed using MS Excel
program. Let y1,y2 and y3 be the quantities of fuel consumed for x1,x2 & x3 t/hr of
steam generated in boilers 1,2 & 3.
Then the total quantity of fuel consumed = y1 + y2 + y3 , which has to be
minimized. Regression equations for fuel consumption as a function of Boiler loads for
boilers 1,2 and 3 ( using Excel Spreadsheet ) are
y1 = 53.52381* x1 + 2196.9
y2=56.911678*x2 + 2124.82
y3 = 47.496883*x3+3191.98
Total fuel consumed at steam production rates x1,x2 & x3 are
F = 53.52381* x1+56.911678*x2+47.496883*x3+ 7513.7
The ultimate linear programming model for minimizing fuel consumption is given by
Minimize 53.52381* x1+56.911678*x2+47.496883*x3+ 7513.7
Subject to
x1+x2+x3 =430
x1<=120
x2 <=180
x3<= 210
x1>=0
x2>=0
x3>=0
40 Energy efficiency Optimization for Power plants. Faculty : Dr.G.G.Rajan
Solution (From Lingo Program)
LP OPTIMUM FOUND AT STEP 2
OBJECTIVE FUNCTION VALUE = 22088.37 kg fuel
VARIABLE VALUE REDUCED COST
X1 120.000000 0.000000
X2 100.000000 0.000000
X3 210.000000 0.000000
NO. ITERATIONS = 2
As per existing operating pattern presented above, the fuel consumption is calculated as
29952.3 kg/hr which is higher than the optimum value by 7863.9 kg/hr. i.e. 35.6 % than
the optimum consumption. This example shows how operation of boilers can be
optimized to minimize fuel consumption using operations research techniques.
Case study 1
A power generation unit has five nos of power boilers B1,B2,B3,B4 & B5 whose
operating parameters are given in the following table.
Boiler
code
Steam
Generation
capacity t/hr
Efficiency at
full load in
%
Steam / fuel
weight ratio at
design load
B1
B2
B3
B4
B5
300
250
200
350
400
87.5
85.0
79.5
91.0
92.5
11.5
11.0
10.5
12.0
12.5
41 Energy efficiency Optimization for Power plants. Faculty : Dr.G.G.Rajan
Steam pressure : 100 kg/cm2g : Steam temperature = 450 oC. Fuel used = FO:
Calorific value of fuel (NCV) = 10000 kcal/kg.
Total steam demand to generate power in two identical turbines is 1300 t/hr.
Determine the steam load on each boiler so that the overall efficiency of boiler is
maximum and fuel consumption is minimum.
Logical programming:
Though this is basically a linear programming problem, this can be done even by
mere logical data analysis. In this case boiler efficiencies are linear and proportional to
actual steam production. Hence, it is logical to load the boiler which gives maximum
efficiency. In this case boiler B5 shows an efficiency of 92.5 % at full load of 400 t/hr.
Hence B5 load is taken as 400. From steam demand point of view, the balance to be
generated out of other boilers is 900 ton/hr and boilers available are B1,B2,B3 and B4.
Out of these four boilers, B4 shows an efficiency of 91 % at full load. Hence B4 load is
taken as 350 t/hr. Other boilers left out are B1,B2 and B3 and the balance steam demand
is 550 tons/hr. Next boiler in the efficiency hierarchy is B1 which may be loaded to 300
t/hr. Balance demand of 250 t/hr may be met by boiler B2. B3 will be just idling.. This is
the simplest logical method followed by plant managers in real life situations .
Boilercode
Steam Generation
capacity t/hr
Efficiency at full load in %
Steam / fuel weight ratio at design load
B5
B4
B1
B2
400
350
300
250
92.5
91.0
87.5
85.0
12.5
12.0
11.5
11.0
42 Energy efficiency Optimization for Power plants. Faculty : Dr.G.G.Rajan
In this case, the overall efficiency of the boiler system shall be
t = (400 x 0.925 + 350 x 0.91 + 300 x 0.875 + 250 x 0.85) / 1300
= 89.5
Fuel consumption = (400/12.5) + (350/12.0)+ (300/11.5) + (250/11.0)
= 32 + 29.17 + 26.09 + 22.73 = 110.53
Average steam / fuel = 11.76
When the steam demand changes, this exercise has to be repeated time and again .
Manual calculation being a cumbersome process, this may be programmed using
operations research techniques as shown in example 1 .
This problem may be quantitatively stated and solved as explained below.
Let x1,x2,x3,x4 and x5 be the steam loads. The main objective is to meet the steam
demand and maximize boiler efficiency and minimize fuel consumption. Hence a
mathematical relationship may be developed between steam generated quantity and fuel
consumed.
Efficiency of Boilers 1,2,3,4 & 5 at their respective loads x1,x2,x3,x4 & x5 will be
0.0029167 x1, 0.0034 x2, 0.003975 x3, 0.0026 x4 and 0.0023125 x5 respectively.
e.g. Calculation for boiler 1. : 1 = (x1/300) x 0.875 = 0.0029167
Efficiency for other boilers is calculated as shown for boiler 1.
Overall efficiency of boiler system = t = x1 * 1 + x2 * 2 + x3 * 3 + x4 * 4
x1 + x2 + x3 + x4
Substituting these values of efficiency function in the above equation, we get
t = 0.0029167 (x1)2 + 0.0034 (x2)2 + 0.003975(x3)2 + 0.0026(x4) 2 + 0.0023125 (x5)2
x1 + x2 + x3 + x4
43 Energy efficiency Optimization for Power plants. Faculty : Dr.G.G.Rajan
where t is the overall efficiency of the boiler system.
The problem may now be stated as
Maximize
t = 0.0029167 (x1)2 + 0.0034 (x2)2 + 0.003975(x3)2 + 0.0026(x4) 2 + 0.0023125 (x5)2
x1 + x2 + x3 + x4
Subject to x1+x2+x3 + x4 + x5 = 1300 ( steam demand )
x1 < = 300 (boiler 1 load limit)
x2 < = 250 (boiler 2 load limit)
x3 < = 200 (boiler 3 load limit)
x4 < = 350 (boiler 4 load limit)
x5 < = 400 (boiler 5 load limit)
x1,x2,x3,x4 , x5 > = 0
This represents a non linear objective function with constraints given as above. To
solve this model using Lindo, the non linear equation must be converted into a linear
equation. Solving this problem involves a number of complicated steps and it is beyond
the scope of the book.
C. Optimization steps
This is referred to as ‘Problem Formulation’. This is the most critical aspect of
any optimization problem to achieve tangible results. Any snag in the problem
formulation, may result in non-feasible / impractical solutions. Extreme care must be
taken in ‘problem formulation’, as this is the most intelligent activity to tackle any
problem.
Following steps are involved in optimizing the performance of any system.
First step is to define the objective / goal we are aiming at.
Examples given below are typical objective functions related to production or
operation of an industry. They are
44 Energy efficiency Optimization for Power plants. Faculty : Dr.G.G.Rajan
Maximizing production quantity
Maximizing operating profits
Minimizing operating costs
Minimizing energy consumption / energy costs
Minimizing emissions level
Loss reduction etc
Step 2: Identify process constraints which have an impact on the objective function.
Typical examples are
Capacity utilization or load factor in boilers, heaters, fans, blowers, turbines etc.
(As capacity utilization increases, specific energy consumption reduces and efficiency
increases up to certain capacity (i.e. design capacity ) and starts dropping down, when
the capacity utilization increases further.)
Run length .(In many equipments, wear and tear results in loss of efficiency
which will result in higher energy consumption. In the case of boilers, heaters, heat
exchangers etc fouling increases with operating period, which retard heat transfer and
decrease equipment efficiency and increase fuel consumption )
Operating severity. This refers to the reaction temperature / pressure / recycle
etc which reduce the energy efficiency of the system. In the case of conversion processes,
the objective is to increase conversion at the expense of energy / operating costs.
Feed quality / composition : In petroleum refining, petrochemical, fertilizer
plants etc feed quality / composition plays an important part in operating profit,
production rate etc.
Step 3: Establish a mathematical model between these variables and the objective
function. If these variables are linearly connected, the problem may be solved by Linear
Programming methods else it has to be solved by NLP / Parametric programming
methods.
45 Energy efficiency Optimization for Power plants. Faculty : Dr.G.G.Rajan
In the boiler loading example , relationship between fuel consumption and steam
generation quantity was established using linear regression models for all the three
boilers. Using these models, total fuel consumption was established by summing up these
relationship to get an objective function equation. In the case study, where five boilers
were considered, a non linear model was developed to determine overall boiler efficiency
by individual efficiency models. In this case the objective function was to maximize
overall efficiency of the system of five operating boilers.
Step 4: Identify the constraints which impede / control the objective function. In the
example dealing with three boilers, the constraint was basically the maximum steam
generating capacity. There could be any number of constraints which will be incorporated
in the LP model.
Step 5: When once all the above steps have been applied and problem formulated
successfully, the solution may be arrived at using Operations Research tools such as
Linear / Non Linear / Parametric programming methods.
C. Common Industrial optimization problems
Single variable problem: Variables used in these problems are called
‘manipulable / endogenous ‘ variables as these may be adjusted by the user at his will.
For example, the feed rate in a heater or steam production rate from a boiler may be
varied physically by the operations staff .
What is not within his control is the heater / boiler efficiency. All the other
parameters remaining constant, heater / boiler / pump / turbine efficiency etc varies with
the load based on the load characteristic curve of the equipment which is design specific.
A typical load vs efficiency curve is shown in fig 2.02. It may be noted from the figure ,
that the equipment efficiency is maximum (87.5%) at 100 % capacity utilization.
Many designers often incorporate addition capacity of 5 to 10 % to take care of
any deficiencies. Under these conditions, the maximum efficiency point will slightly shift
46 Energy efficiency Optimization for Power plants. Faculty : Dr.G.G.Rajan
to the right than what is shown in the figure. Similarly when the equipment is under
designed, the peak efficiency shall shift to the left. This aspect is shown in fig 2.03.
80
81
82
83
84
85
86
87
88
70 80 90 100 110 120
load %
effic
ienc
y
blreff %
Fig 2.02. Efficiency vs Capacity utilization – Generic Model
47 Energy efficiency Optimization for Power plants. Faculty : Dr.G.G.Rajan
Above information may be represented in mathematical form to arrive at
quantitative decisions. Since the efficiency vs load shows a bell shaped curve, the model
of the form (quadratic model) ax2+bx+c or a cubic equation shall be more appropriate.
The same data may be used to develop these models. For developing these regression
models, principle of least square is used and the coefficients arrived at. When once the
quantitative models are developed, it is easy to find out the peak efficiency point.
76
78
80
82
84
86
88
90
70 80 90 100 110 120
load %
effic
ienc
y %
10% under normal 10% over
Fig 2.03. Impact of Design on peak efficiency point
48 Energy efficiency Optimization for Power plants. Faculty : Dr.G.G.Rajan
Models are of the form
Eff % = -6.198286E-03 * x2 + 1.277158 * x + 20.74732 (I)
Eff % = -2.266988E-04 * x3 + 5.842239E-02 * x2 - 4.748416 * x + 204.3378 (II)
Standard error of model I and II are 0.7594 and 0.3769 respectively.
x is the load % of the equipment.
78
80
82
84
86
88
90
load %
effic
ienc
y %
actual quadratic polynomial
actual 80.5 82.1 85.0 87.6 86.5 84.3
quadratic 79.8 83.3 85.5 86.5 86.2 84.8
polynomial 80.5 82.3 84.9 87.0 87.2 84.1
70 80 90 100 110 120
Fig 2.04. Efficiency vs Load % models
49 Energy efficiency Optimization for Power plants. Faculty : Dr.G.G.Rajan
D. Time dependant optimization models
Let us consider the case of a process heater whose efficiency varies with load %
as shown in the previous example. The same heater will perform differently with passage
of time say 1 yr, 1½ yrs, 2 yrs etc. Though the efficiency pattern may remain the same in
all these load conditions, observed efficiency for the same load will be somewhat lower
with passage of time. Hence necessary run length factor will have to be incorporated in
the model. A typical case is shown below and in fig 2.05.
Above table shows the behavioral pattern of the heater with passage of time
called ‘on-stream hours’. It is possible to predict what will be the behavioral pattern after
certain passage of time for taking corrective action. For this purpose, a two variable
model is built for the heater with load and operating hours as two independent variables.
This is a very important method for heat exchangers / coolers / condensers, as they
develop fouling with time and even after de-scaling, the original heat transfer rate is
seldom achieved. Above data when converted into a model gives,
Efficiency % = 0.1416667 * load % - 0.1101426 *op. month + 71.2722
This model may be used to predict the efficiency at any load over a reasonable
operating month. For example, the predicted heater efficiency by 35th month at 100 %
load will be of the order of 81.58 % .
Load % Base case After 1 year
After 2 years
70 80 79 78.0 80 82 80 78.0 90 85 83 81.5 100 88 87 85.5 110 85 82 81.0
50 Energy efficiency Optimization for Power plants. Faculty : Dr.G.G.Rajan
Fig 2.05. Efficiency deterioration with on stream hours
51 Energy efficiency Optimization for Power plants. Faculty : Dr.G.G.Rajan
E. Linear Programming (LP) Problems
We have shown some examples related to boiler efficiency optimization, which
represented a linear programming (LP) problem, in which the objective and the
constraints were linear functions of the decision variables. Typical example of a linear
function is:
75 x1 + 50 x2 + 35 x3
where x1, x2 and x3 are decision variables. The variables are multiplied by coefficients
(75, 50 and 35 above) that are constant in the optimization problem. They may be
computed in Excel worksheet. Linear programming problems are intrinsically easier to
solve than nonlinear (NLP) problems. In an NLP there may be more than one feasible
region and the optimal solution might be found at any point within any such region. In
contrast, an LP has at most one feasible region with “flat faces”. Some simpler problems
may be solved graphically. A typical example related to a heater is given below.
Case study 2
A fired heater having an operating capacity of 500 t/hr of feed uses fuel gas and
fuel oil to heat the feed to the outlet temperature of 345 oC. The calorific value of fuel oil
is 9900 kcal / kg and that of fuel gas is 10800 kcal/kg. Cost of fuel oil is 100 us$ / ton
while that of fuel gas is 60 us$/ton. Maximum Fired duty of heater is 100 mmkcal /hr.
From radiation temperature point of view, the heater should have minimum 65% oil
firing and maximum 90%. Atomizing steam is 15% of oil consumed and the cost of
steam is 45 us$/ton. Availability of fuel gas is limited to 4.0 t/hr maximum. Find the
optimum fuel mix that will minimize the total operating cost of the heater.
Formulating the problem:
Let x1 and x2 be the quantity of fuel oil and fuel gas fired in the heater to meet the heat
duty.
Total cost of fuel = 100 * x1 + 65 * x2 (I)
52 Energy efficiency Optimization for Power plants. Faculty : Dr.G.G.Rajan
Cost of steam = 0.15 * x1 * 45 (II)
Total cost = 100 * x1 + 65 * x2 + 0.15 * x1 * 45
= 106.75 * x1 + 65 * x2 (III)
Objective function is to minimize fuel cost. i.e. to minimize 106.75 * x1 + 65 * x2
The total calorific value of the fuel mix supplied must meet the required fired duty. Since
the calorific value of Fuel Oil and fuel gas are 9900 and 10800 kcal/ton respectively, the
heat duty equation is formulated as
x1 * 1000 * 9900 + x2 * 1000 * 10800 = 100 * 1000000
i.e. x1 + 1.0909091 * x2 = 10.10101 (IV)
Fuel oil fired should be minimum 65 %
x1 >= 0.65 * 100 * 1000000 / (9900 * 1000) tons
i.e. x1 >= 6.565656 (V)
Fuel oil fired should be maximum 90 %
i.e. x1 <= 0.90 * 100 * 1000000 / (9900 * 1000)
i.e. x1 <= 9.09091 tons. (VI)
Fuel gas availability constraint is 4.0 t/hr.
x2 < = 4.0 (VII)
The final problem may now be written as
Minimize 106.75 * x1 + 65 * x2 (I)
Subject to
x1 + 1.0909091 * x2 = 10.10101 (II)
x1 >= 6.565656 (III)
x1 <= 9.09091 tons. (IV)
x2 < = 4.0 (V)
53 Energy efficiency Optimization for Power plants. Faculty : Dr.G.G.Rajan
This is comparatively a very simple problem that may be visualized in graphical form and
may be solved by logical programming as explained below.
Since the fuel oil to be used is minimum 6.565656 tons, x2 may be calculated using
equation II.
i.e. 1.0909091 * x2 = 10.10101 – 6.565656 = 3.535354
i.e. x2 = 3.24074 tons.
This meets all the constraints.
Hence total fuel cost ( objective function) is 911.531878 us$ .
Surplus gas available due to the optimal fuel mix is = 4 –3.24074 = 0.75926 tons
Case b: The operations manager decides to use the total fuel gas in the heater by making
some operational adjustments so that the minimum oil firing reduces to 50 %. What
impact does it make on the operating cost ?
In this case constraint III may be modified as
x1 >= 0.5*100*1000000/(9800*1000)
x1 >= 5.102041 tons. (III)
Since fuel gas availability is 4.0 tons / hr and the entire fuel gas will be consumed, the oil
required to meet the heat duty may be calculated by equation (II) as given below.
x1 + 1.0909091 * x2 = 10.10101
i.e. x1 + 1.0909091 * 4.00 = 10.10101
i.e. x1 = 10.10101 - 4.3636364 = 5.7373736
This meets modified constraint III and all the other constraints.
Total fuel cost in this case is = 106.75 * x1 + 65 * x2 = 106.75 * 5.7373736 + 65 * 4.00
= 872.4646318
The reduction in fuel cost due to the revised fuel mix is
= 911.531878 -872.4646318
= 39.0672462 us$/hr = 4.2858 % reduction.
54 Energy efficiency Optimization for Power plants. Faculty : Dr.G.G.Rajan
Based on 8000 operating hours, total annual savings due to the revised fuel mix works
out to 312537.9 us$ / annum for a single heater. This is fairly a simple problem where
only one heater was considered with minimum constraints.
Case study 3. Two heaters in parallel service
Two heaters having an operating capacity of 500 t/hr and 200 t/hr of feed are
operating in parallel. Both heaters use fuel gas and fuel oil to heat the feed to the outlet
temperature of 345 oC. The calorific value of fuel oil is 9900 kcal / kg and that of fuel
gas is 10800 kcal/kg. Cost of fuel oil is 100 us$ / ton while that of fuel gas is 60 us$/ton.
Maximum Fired duty of heater1 is 100 mmkcal /hr, while that of heater2 is 20
mmkcal/hr. Heater1 should have minimum 5.5 t/hr oil firing and maximum of 9.10 t/hr.
Heater2 should have minimum 1.0 t/hr oil firing. Atomizing steam is 15% of oil
consumed and the cost of steam is 45 us$/ton. Availability of fuel gas is limited to 5.0
t/hr maximum. Find the optimum fuel mix that will minimize the total operating cost of
the heater ?
Formulating the problem:
Let x1 & x2 be the quantities of fuel oil used in heaters 1 & 2 respectively. From
the fired duty, fuel gas consumed may be back calculated as shown below.
The objective function is to minimize fuel cost in the combined operation of heaters
which is stated as
Minimize x1*100 + x1*0.15*45 + x2*100 + 0.15*x2*45 + 60 * (100-9.8 x1)/10.8 + 60
* (20-9.8x2)/10.8
= 106.75 * x1 + 106.75 * x2 + 555.55 - 54.4444* x1 + 111.1111 - 54.4444 * x2
= 52.3056 * x1 + 52.3056 * x2 + 666.6611
Minimize 52.3056 * x1 + 52.3056 * x2 + 666.6611 (I)
ST
x1 >= 6.565656 (II)
55 Energy efficiency Optimization for Power plants. Faculty : Dr.G.G.Rajan
x1 <= 9.09091 tons. (III)
x2 >= 1.0204 (IV)
Since the coefficients in the objective function are equal, only constraints II & IV are
important in this case.
Maximum amount of gas that may be consumed in heater 1 is = 3.3015 tons.
Remaining fuel gas available is = 6-3.3015 = 2.6985 tons
Maximum amount of gas that may be consumed in heater 2 is = 0.92592 tons
Total gas consumed = 4.2274 tons.
Gas left unutilized = 0.7726 tons
This is again based on logical programming. This problem may be solved by
using LP models considering both liquid and gas fuels individually . In this case, the
objective function will undergo a change.
Item Heater 1 Heater II Fuel oil in tons
x1 x2
Equivalentheat inmmkcal
x1*9800*1000/1000000=9.8 x1 = 9.8 x2
Balanceheat load
(100-9.8 x1) (20-9.8x2)
Eqvt fuel gas in tons
(100-9.8 x1)/10.8 (20-9.8x2)/10.8
56 Energy efficiency Optimization for Power plants. Faculty : Dr.G.G.Rajan
F. LP method of solving the two heater problem :
Let x1 & x2 be the quantities of fuel oil used in heaters 1 & 2 respectively and x3
and x4 quantity of fuel gas used in the heaters 1 & 2 respectively.
The objective function is to minimize total fuel costs meeting fired duty and % fo
constraints . Considering the atomizing steam cost and other constraints the problem is
formulated as shown below.
Output from Lingo 8 program.
MIN = 106.75 * x1 + 106.75 * x2 + 60 * x3 + 60 * x4 ;
x3 + x4 = 5 ;
9.9*x1+10.8*x3 = 100;
9.9*x2+10.8*x4 = 20;
x1 >= 5.5;
x1 <= 9.09 ;
x2 >= 1.00;
x3 >= 0;
x4 >=0;
Global optimal solution found at iteration: 2
Objective value: 1011.667
Variable Value Reduced Cost
X1 5.500000 0.000000
X2 1.166667 0.000000
X3 4.217593 0.000000
X4 0.782407 0.000000
Note : In this case, entire gas has been consumed without any excess quantity. These
examples indicate, how practical optimization helps in reducing the fuel cost at micro
level. The success of this exercise is based on effective problem formulation especially in
defining the objective function and the constraints.
57 Energy efficiency Optimization for Power plants. Faculty : Dr.G.G.Rajan
G. Complex Fuel Mix Problem :
In real life situations, process industries import a variety of fuel oils for their
consumption in process heaters / boilers to manufacture the product. As mentioned in
chapter 1, the energy input such as fuel, steam, power etc varies with the type of industry
and the design factors. Let us consider a complex fuel mix problem, with more number of
liquid fuels with varying sulfur levels.
Cost of sulfurous fuels are comparatively cheaper than clean fuel. Environmental
regulations limit the SO2 emission level from each heater / boiler .High sulfur fuel will
emit more SO2 than a clean fuel. There is an economic balance between use of sulfurous
fuel versus reduction in fuel cost.
Using Linear Programming models, it is possible to determine the maximum
amount of Sulfurous fuel that can be used by a fired heater / boiler such that the
emissions are contained within allowable limits at minimum fuel cost. This is a typical
environmental problem faced by process industries.
H.Impact of process modification in the boiler.
The boiler operating cost can be lowered down , by incorporating certain modifications in
the boiler flue gas sections. From the LP model, it is obvious that the constraints holding
the key to using low cost fuels are SO2 and SPM emission levels. At this stage, it is
possible to incorporate certain cost effective modifications. For removal of SO2 from the
flue gas, a Sulfur Guard may be provided before the air pre heater. This is nothing but a
Zinc wire mesh which is located in a convenient place for easy removal and replacement.
This will react with SO2 present in the flue gas by chemical reaction and is converted
into ZnSO4. After certain passage of time, the zinc mesh may be replaced, when the
reaction becomes ineffective. For removal of SPM, bag filters / or other type of filtering
material may be used. Let us visualize the scenario with S guard & filter modifications as
proposed.
58 Energy efficiency Optimization for Power plants. Faculty : Dr.G.G.Rajan
Let C be the cost of modifications in US$ and project life at Y years. Incremental cost
due to the project on hourly basis ( without considering interest rate ) is given by
Incremental Cost / year = C/Y
Assuming 8000 operating hrs / year, incremental operating cost / hr = (C/8000Y)
Under these conditions, the boiler can generate the entire steam using fuel x3.
Hourly saving due to switching over to 100% of EHFO = 4226.7 – 4165.9 = 60.8
US$.
Let us assume the investment cost at 2,50,000 us$.
Extra cost incurred / hr = 10.42 US$.
Hence the effective net savings / hr = 60.8 – 10.42 US$ = 50.53 US$ /hr
Pay back period = 2,50,000 / (60.8 * 8000) = 0.51 year
This is just a hypothetical case. If the management has an investment policy for projects
with longer pay backs, the investment limit may be doubled or tripled.
It is now obvious from the above practical examples that an LP Solver needs to consider
many fewer points than an NLP Solver, and it is always possible to visualize (subject to
the limitations of finite precision computer arithmetic) that some LP problem
(i) may have no feasible solution
(ii) have an unbounded objective, or may have a globally optimal solution
59 Energy efficiency Optimization for Power plants. Faculty : Dr.G.G.Rajan
I. Quadratic Programming (QP) Problems
A quadratic programming (QP) problem has an objective which is a quadratic function of
the decision variables, and constraints are all linear functions of the variables. A typical
example of a quadratic function is:
2 x12 + 3 x22 + 4 x1 x2
where x1, x2 and x3 are decision variables.
A widely used QP problem is the Markowitz mean-variance portfolio optimization
problem using a number of normal equations. The linear constraints specify a lower/
upper bound for profitability function.
QP problems, like LP problems, have only one feasible region with "flat faces" on its
surface (due to the linear constraints), but the optimal solution may be found anywhere
within the region or on its surface.
Quadratic Programming Problem – Example
Two boilers B1\and B2 are designed to operate at maximum load of 500 and 750 t/hr.
The minimum turn down ratio of B1\ & B2 are 65% and 55 % respectively. Boiler
efficiency characteristics as a function of load for the two boilers are shown in fig 2.06.
Find the optimal load on the two boilers when steam demand is 800, 900 and 1000 t/hr
such that the overall efficiency is maximum.
This may be represented by
Efficiency of boiler 1 is of the form e1 = a1* x12 + b1* x1 + c1
Efficiency of boiler 2 is of the form e2 = a2* x22 + b2* x2 + c2
Over all efficiency of the boiler = (e1*x1*c1 + e2*x2*c2) / (x1*c1+x2*c2)
e1 = -6.000395E-03 * x12 + 1.159349 * x1 + 23.45476
S.E of model : 0.3543
e2 = -6.198286E-03 * x22 + 1.277158 * x2 + 20.74732
S.E of model : 0.7594
Where x1 & x2 are % of load on design, c1& c2 are boiler capacities
60 Energy efficiency Optimization for Power plants. Faculty : Dr.G.G.Rajan
The QP model may now be written as
Maximize (e1*x1*c1 + e2*x2*c2) / (x1*c1+x2*c2)
Subject to (x1*c1 + x2*c2)/100 = 800
x1 >=65
x2 >= 55
65
70
75
80
85
90
Load %
blr
eff%
LCBLR HCBLR
LCBLR 66.4 71.4 75.2 77.8 79.2 79.4 78.4HCBLR 69.1 75.1 79.8 83.3 85.5 86.5 86.2
50 60 70 80 90 100 110
Fig 2.06. Overall efficiency optimization – two boilers
61 Energy efficiency Optimization for Power plants. Faculty : Dr.G.G.Rajan
Lingo Program output:
Max = ((-.006000395 * x1^3 + 1.159349 * x1 ^ 2 + 23.45476*x1)*500+750*( -
.006198286 * x2^3 + 1.277158 * x2 ^ 2 +20.74732*x2))/ (500*x1+750*x2) ;
500*x1 +750* x2 = 80000;
x1 >=65; x2 >= 55;
Local optimal solution found at iteration: 17
Objective value: 75.42680
Variable Value Reduced Cost
X1 65.00000 0.000000
X2 63.33333 0.000000
Row Slack or Surplus Dual Price
1 75.42680 1.000000
2 0.000000 0.4063477E-03
3 0.000000 -0.6137188E-01
4 8.333333 0.000000
i.e. Optimum load % x1 & x2 are = 65 and 63.3333%
i.e. 325 tons /hr and 475 tons / hr respectively. Overall boiler efficiency at this load
distribution will be 75.43 %. The same was solved by Logical Programming the output of
which is shown in the table below.
load t/hr load t/hr x1 x2 eff1 eff2 overall eff
325.0 475.0 65.0 63.3 73.46 76.77 75.43
330.0 470.0 66.0 62.7 73.83 76.44 75.37
340.0 460.0 68.0 61.3 74.54 75.76 75.25
350.0 450.0 70.0 60.0 75.21 75.06 75.13
360.0 440.0 72.0 58.7 75.82 74.34 75.01
370.0 430.0 74.0 57.3 76.39 73.60 74.89
380.0 420.0 76.0 56.0 76.91 72.83 74.77
387.5 412.5 77.5 55.0 77.26 72.24 74.67
62 Energy efficiency Optimization for Power plants. Faculty : Dr.G.G.Rajan
The problem will become more complex when more and more variables are involved in
the QP. In such cases, it is a good practice to develop an Evolutionary Operations
Research Model for the required objective function incorporating all the related variables
in the model. Then the problem may be solved by any one of the solvers like Lindo,
Lingo, Excel Solver etc.
J. Evolutionary Operations Research Models for Optimization
Evolutionary Operation Research Technique or (EORT in short) involves very
systematic small changes in process variables during the operation of the process. The
results of previous small changes may be used to suggest further changes so as to
approach the optimum operating conditions by a series of small steps. Care is taken to
ensure that these individual changes in process parameters do not upset the process nor
produce any undesirable outcomes.
Evolutionary Operations Research technique may be used to identify the
combination of multi variables to enhance the response surface of any operation which in
turn improves the operational objectives and also the productivity of the complex system
considerably.
The basic concept underlying EORT is that a smooth response surface exists for a
set of variables, which ultimately tend to converge to a single optimum. This is the
principle of Advanced Control System where the impact of minor variations of process
parameters are used to vary other parameters to maximize or minimize the objective
function. ( e.g maximize energy efficiency of the equipment / total system ) .EORT
models are based on actual process variables which are dynamic in nature and fluctuate
due to exogenous or endogenous factors.
Since the effect of these manipulable variables form the inputs to the model, the
outcomes will naturally be far more precise and accurate. EORT models may be used for
carrying out the most needed sensitivity analysis of the dynamic system to identify the
output trend for a set of new variables and/or operating conditions. EORT models may
63 Energy efficiency Optimization for Power plants. Faculty : Dr.G.G.Rajan
also be used to determine and monitor the system efficiency as a whole and also the sub
system constituting them. Hence corrective actions may be taken more precisely & in
time.
EORT models are real life operating models and combines the expertise of the
operations group and the system response to arrive at the right decisions compatible with
the system.
EORT models are deterministic in nature as the variables and their inter
relationship are evaluated in real life situations by observed values and performance
rather than theoretical / design mode models which suffer from many inherent
assumptions.
EORT models are system specific and cannot be generalized to all similar
situations. For developing a generalized model all variables constituting the total macro
system will have to be considered. EORT generates a set of Decision Models based on
LP / NLP algorithms and serve as a very powerful, effective and result-oriented Decision
Support System (DSS in short).
These quantitative Decision Models are logical abstraction of reality under which
the total complex system operates. While simulation models are based on certain
assumptions and hypothesis, which in practice may not be totally true or valid, EORT
models are based on observed facts and figures. All constraints experienced in real life
situations are reflected in the model output.
This type of DSS serves as a conduit for creating, revising, reviewing and
checking the real performance of the total system or subsystems which are constituents of
the main system. Also it is possible to achieve the operational objectives more easily than
one can visualize.
64 Energy efficiency Optimization for Power plants. Faculty : Dr.G.G.Rajan
K. EORT Modeling Methodology
Following steps are to be followed meticulously to develop a very reliable and valid
model for the energy industry.
Defining the expected outcome/ objective of the model
Outlining the general process & data collection
Identifying the manipulable and nonmanipulable variables either by observed
data, analytical data or a combination of both.
Developing basic Model with the identified variables which affect the objective
function, and deleting the least effective variables by model revalidation.
Sensitivity analysis and model validation by manipulating the variables and
comparing the observed output with model output.
Checking model validity under observed conditions for a stipulated range of
variables.
No validation will be deemed necessary if the deviation between observed and
model outputs are within statistical limits.
When once the last stage has been reached, the model will be accepted as valid
and will form a basic DSS tool, by which the actual performance of the system may be
evaluated and compared and the reason for any discrepancy may be identified for
remedial action.
If the variation is explainable, the Decision Maker may correct the situation
immediately. If the variation is not explainable or much higher than the stipulated limits,
the situation must be taken as out of control and warrants thorough investigation. Hence
EORT models are capable of giving the warning signals at right time and could easily
avert inefficiency or even disaster in some cases.
65 Energy efficiency Optimization for Power plants. Faculty : Dr.G.G.Rajan
L. Application of EORT Models
From experience it has been found that EORT Models have unlimited application
areas in the process industry.
Typical applications relate to
Cooler / Condenser performance prediction
Cooling water quality monitoring
Evaluation of inhibitor performance
Equipment deterioration study
Specific Energy consumption of a single process unit or the entire industry.
Some of the examples given here refer to actual observations of some of the industries.
Using EORT Models in the objective function.
Since EORT models combine all the independent variables in the form of an
equation and the outcome is defined, they may be used as the objective function and the
limits of the independent variables shall be the constraints.
Case Study 5 - Energy consumption model using EORT model. This example refers to the Power consumption of a Centrifugal air compressor
based on the observed operating data such as air flow rate, rpm , discharge pressure and
power consumption. Following table gives the details of parameters observed during
actual run of the air compressor. The objective is to minimize power consumption for
certain air flow and required discharge pressure.
Flow nm3/hr RPM Disch Pr. ata
PowerConsumption kwh
70000 5355 5.00 5700 75000 5355 4.95 6100 80000 5355 4.90 6300 85000 5355 4.75 6500
66 Energy efficiency Optimization for Power plants. Faculty : Dr.G.G.Rajan
90000 5355 4.60 6600 95000 5355 3.80 6700 100000 5355 3.20 6750 70000 5610 5.70 6400 75000 5610 5.60 6700 80000 5610 5.45 7000 85000 5610 5.30 7300 90000 5610 5.10 7500 95000 5610 4.80 7700 100000 5610 4.55 7800 105000 5610 3.90 7900 75000 5891 6.40 7600 80000 5891 6.30 8000 85000 5891 6.15 8300 90000 5891 5.95 8700 95000 5891 5.80 8900 100000 5891 5.50 9000 105000 5891 5.25 9100
EORT Model is of the form
KW = 2.864448E-08* (Flow) 0.64983*(RPM) 2.15599*(Pr) 0.16322
The validity of the model is seen from the model output of power consumption
versus observed power consumption in KWH as shown below.
Compressor Performance Model
observed simulated error power kw/h by .Model term
5699.998 5738.24756 -38.25002
6100.002 5991.52393 108.47852
6300.002 6237.79688 62.20567
6499.999 6455.61182 44.38672
6599.998 6664.90674 -64.90868
67 Energy efficiency Optimization for Power plants. Faculty : Dr.G.G.Rajan
6700.002 6691.28564 8.71616
6750.001 6726.73193 23.26927
6400.000 6480.75488 -80.75426
6700.002 6758.36816 -58.36635
7000.000 7016.66162 -16.66237
7300.002 7265.44189 34.55997
7499.998 7493.19873 6.79900
7700.000 7684.71875 15.28089
7799.998 7876.11523 -76.11685
7900.000 7927.84131 -27.84197
5699.998 5738.24756 -38.25002
6100.002 5991.52393 108.47852
6300.002 6237.79688 62.20567
6499.999 6455.61182 44.38672
6599.998 6664.90674 -64.90868
6700.002 6691.28564 8.71616
6750.001 6726.73193 23.26927
6400.000 6480.75488 -80.75426
6700.002 6758.36816 -58.36635
7000.000 7016.66162 -16.66237
7300.002 7265.44189 34.55997
7499.998 7493.19873 6.79900
7700.000 7684.71875 15.28089
7799.998 7876.11523 -76.11685
7900.000 7927.84131 -27.84197
7600.000 7674.84619 -74.84582
8000.001 7983.02832 16.97263
8299.999 8271.21289 28.78612
68 Energy efficiency Optimization for Power plants. Faculty : Dr.G.G.Rajan
8700.000 8538.01855 161.98141
8899.997 8806.54102 93.45620
8999.997 9026.44629 -26.44888
9100.000 9246.72559 -146.72505
Standard Error of the estimate = 71.06937
Error level of this model being less than 1.0%, this may be used for optimizing
compressor performance. Manipulable parameter in this case is RPM. Air flow rate and
Discharge pressure of the compressor are the process requirement. Hence the problem
may be defined as minimize power consumption subject to the required air flow rate
(nm3/hr) and Discharge pressure in ATA. ( Atmosphere absolute). Using the EORT
model, the problem may now be formulated as
69 Energy efficiency Optimization for Power plants. Faculty : Dr.G.G.Rajan
From the above example, the need for modeling the complex system using manipulable
variables and their impact on objective function was shown. It was also shown that this
model can be used in NLP optimization model, using the constraint equations to arrive at
the optimal solution.
Case Study 6
Three centrifugal compressors are used in an oxidation plant for various applications such
as Oxidation process, CO boiler and other process use. Compressor characteristics of the
compressors are given below. The operating range of the compressors are shown below.
KW1 = 2.86E-08* (F1)^0.64983*(RPM1)^ 2.15599*(Pr1)^0.16322
KW2 = 2.89E-08* (F2)^0.65*(RPM2)^ 2.45599*(Pr2)^0.19322
KW3 = 2.88E-08* (F3)^0.65*(RPM3)^ 2.45599*(Pr3)^0.175
EORT Model application in optimization
Min = 2.864448E-08* (Flow)^0.64983*(RPM)^ 2.15599*(Pr)^0.16322; Flow = 95000; Pr = 5.85; RPM <=5800; RPM >= 5300;
Local optimal solution found at iteration: 12
Objective value: 7021.530
Variable Value Reduced Cost
FLOW 95000.00 0.000000
RPM 5300.000 0.000000
PR 5.850000 0.000000
Row Slack or Surplus Dual Price
1 7021.530 -1.000000
2 0.000000 -0.4802949E-01
3 0.000000 -195.9033
4 500.0000 0.000000
5 0.000000 -2.856293
Since the minimum RPM of the compressor is 5355, we shall stipulate this condition in the model , by
modifying the last line in the model as RPM >= 5500. The modified output is given below.
Local optimal solution found at iteration: 12
Objective value: 7605.273
Variable Value Reduced Cost
FLOW 95000.00 0.000000
RPM 5500.000 0.000000
PR 5.850000 0.000000
Note the increase in power consumption by 584 KWH for the revised RPM level. The model shows that the
RPM must be as low as possible for low energy consumption.
70 Energy efficiency Optimization for Power plants. Faculty : Dr.G.G.Rajan
Operating Parameters Range
Flow RPM Pressure (kg/cm2g)
Compressor 1 50000-100000 5000-7500 5.5-8.0
Compressor 2 65000-120000 5000-7500 5.5-8.0
Compressor 3 45000 – 95000 5000-7500 5.5-8.0
Total requirement of air = 195000 nm3/hr
Minimum required Discharge pressure = 7.5 kg/cm2g
What is the optimal operation of these compressors that will minimize the total power
consumption ?
Formulation of the problem:
Min = 2.86E-08* (F1)^0.64983*(RPM1)^ 2.15599*(Pr1)^0.16322+2.89E-08*
(F2)^0.65*(RPM2)^ 2.45599*(Pr2)^0.19322+ 2.88E-08* (F3)^0.65*(RPM3)^
2.45599*(Pr3)^0.175;
F1 + F2 + F3 = 195000;
F1>=50000;
F2>=65000;
F3>=45000;
F1<=100000;
F2<=120000;
F3<=95000;
Pr1 = 7.5;
Pr2 = 7.5;
Pr3 = 7.5;
RPM1 >=5000;
RPM1 <=7500;
RPM2 >=5000;
RPM2 <=7500;
RPM3 >=5000;
71 Energy efficiency Optimization for Power plants. Faculty : Dr.G.G.Rajan
RPM3 <=7500;
Program Output.
Local optimal solution found at iteration: 8
Objective value: 128345.8 KWH
Variable Value Reduced Cost
F1 85000.00 0.000000
RPM1 5000.000 0.000000
PR1 7.500000 0.000000
F2 65000.00 0.000000
RPM2 5000.000 0.000000
PR2 7.500000 0.000000
F3 45000.00 0.000000
RPM3 5000.000 0.000000
PR3 7.500000 0.000000
In the optimization model , EORT model of the three compressors were taken in
the objective function and the operational limits for RPM, Flow and Pressure were used
in the constraint equations. The output shows that the minimum power consumption will
be 128345.8 KWH.
In this problem, it is formulated in such a way that all the three compressors will
be running on load. This is to take care of any eventualities of compressor failure which
may jeopardize the process. For a two compressor operation, the formulation has to be
modified and program rerun to get the optimum results. Since the compressed air demand
is just 195000 nm3/hr, the same can be met by a combination of compressor 1 & 2 or
compressor 2&3 keeping the other idle. Compressor 1 & 3 option is left out as there is no
cushion in air capacity. The output for these 2 combinations is shown below.
72 Energy efficiency Optimization for Power plants. Faculty : Dr.G.G.Rajan
Compressor 1 & 2 in operation. Compressor 3 idle:
Local optimal solution found at iteration: 8
Objective value: 95809.07 kwh
Variable Value Reduced Cost
F1 100000.0 0.000000
RPM1 5000.000 0.000000
PR1 7.500000 0.000000
F2 95000.00 0.000000
RPM2 5000.000 0.000000
PR2 7.500000 0.000000
F3 0.000000 0.000000
RPM3 5000.000 0.000000
PR3 7.500000 0.000000
Note this operation mode results in a power saving of 32536.73 kwh i.e. equivalent to
25.35 % on base case operation. This operation has a cushion of 11.4 % on air supply. If
the demand exceeds this supply, the process will have a set back.
Let us now consider the next option of running compressor 2 & 3 .
Local optimal solution found at iteration: 8
Objective value: 177214.7 kwh
Variable Value Reduced Cost
F1 0.000000 0.000000
RPM1 5000.000 0.000000
PR1 7.500000 0.000000
F2 120000.0 0.000000
RPM2 5000.000 0.000000
PR2 7.500000 0.000000
F3 75000.00 0.000000
RPM3 5000.000 0.000000
73 Energy efficiency Optimization for Power plants. Faculty : Dr.G.G.Rajan
PR3 7.500000 0.000000
This option shows an increase of 48868.9 kwh for the same demand of 195000
nm3/hr. i.e 38% higher than the three compressor operation.
Hence this is not an economically viable option. Consolidated operational
analysis of these three options related to air compressor is given below. From this it is
obvious that Optimization of operational parameters can yield energy savings with
substantial reduction in operating costs. In complex situations, there is no substitute for
quantitative operational analysis involving LP/NLP/EORT/QP techniques.
The most important property of QP's quadratic objective is that it has an
algebraic property of definiteness. The "best" quadratics are positive definite (in a
minimization problem) or negative definite (in a maximization problem). You can picture
the graph of these functions as having a "round bowl" shape with a single bottom (or
top). Refer to figs 2.05 to 2.06. ( maximization problem)
In minimization problems, the curve will be bell shaped. As the value of
independent variable increases, the vale of objective function will become lower, reach a
minimum point and start raising again.
Option Power Consumption
Kwh
Remarks
All 3 compressors in operation
128345.8 Base case
Compressors 1 & 2 only
95809.07 Lowest energy
Compressors 2 & 3 only
177214.7 Highest energy
74 Energy efficiency Optimization for Power plants. Faculty : Dr.G.G.Rajan
Solving LP and QP Problems
LP problems are usually solved via the Simplex method. This method, originally
developed by Dantzig in 1948, has been dramatically enhanced in the last decade, using
advanced methods from numerical linear algebra. This has made it possible to solve LP
problems with up to hundreds of thousands -- sometimes millions -- of decision variables
and constraints.
An alternative to the Simplex method, called the Interior Point or Newton-Barrier
method, was developed by Karmarkar in 1984. Also in the last decade, this method has
been dramatically enhanced with advanced linear algebra methods so that it is often
competitive with the Simplex method, especially on very large problems.
A faster and more reliable way to solve a QP problem is to use an extension of the
Simplex method or an extension of the Interior Point or Barrier method.
Nonsmooth Optimization (NSP)
The most difficult type of optimization problem is a non smooth problem (NSP)
which has multiple feasible regions and multiple optimal points within each region. In
most of these problems, it is impractical to enumerate all the possible solutions and pick
the best one. These techniques rely on some sort of random sampling of possible
solutions. Such methods are non deterministic or stochastic -- they may yield different
solutions on different runs, even when started from the same point on the same model,
depending on which points are randomly sampled.
Solving NSP Problems
Genetic or Evolutionary Algorithms as shown in the previous examples a way to
find "good" solutions to non smooth optimization problems. (In a genetic algorithm the
problem is encoded in a series of bit strings that are manipulated by the algorithm; in an
"evolutionary algorithm," the decision variables and problem functions are used directly.
Most commercial Solver products are based on evolutionary algorithms.) These
algorithms maintain a population of candidate solutions, rather than a single best solution
75 Energy efficiency Optimization for Power plants. Faculty : Dr.G.G.Rajan
so far. From existing candidate solutions, they generate new solutions through either
random mutation of single points or crossover or recombination of two or more existing
points. The population is then subject to selection that tends to eliminate the worst
candidate solutions and keep the best ones. This process is repeated, generating better and
better solutions; however, there is no way for these methods to determine that a given
solution is truly optimal. This method uses memory of past search results to guide the
direction and intensity of future searches. These methods generate successively better
solutions, but as with genetic and evolutionary algorithms, there is no way for these
methods to determine that a given solution is truly optimal.
76 Energy efficiency Optimization for Power plants. Faculty : Dr.G.G.Rajan
L. Simplex method :
Simplex method is the most basic and logical method of solving an LP problem.
In this method all the relationship are converted into equations as explained below.
Linear programming is concerned with solutions to simultaneous linear equations. These
equations are developed on the basis of restrictions on the variables. These restrictions
are always expressed as inequalities. The first step is to convert them into equations.This
is done by incorporating a new variable known as slack variable. For example if the
inequality is of the form a1x1+a2x2+a3x3 < = b , this may be converted into
a1x1+a2x2+a3x3+S1 = b. The slack variable will take the value , which will satisfy the
equation. If the inequality is of the form a1x1+a2x2+a3x3 > = b, this will be converted
into an equation by subtracting a slack variable as given by a1x1+a2x2+a3x3 – S1= b.
This step is carried out for all the constraints, so that only linear equations shall
appear in the problem. For example let us consider a coal fired boiler problem as
explained below.
Case study:
A coal boiler produces 750 t/hr of steam at 100 kg/cm2 pressure and 500 oC. The
boiler uses three grades of coal c1,c2 and c3 with the following specifications.
Coal Sulfur Phosphorous Ash Net CV Wt % Wt % Wt % kcal/kg
c1 3.8 0.02 3.0 4000
c2 1.9 0.01 2.0 4500
c3 1.0 0.005 5.0 4900
Presence of excess sulfur SO2 emissions and based on the environmental
regulations, S in the coal mix should not exceed 1.8% by weight. Phosphorous % is
restricted to 0.015 % by weight. Ash content should not exceed 3.0 %. Fuel efficiency of
the boiler is 86.0, 87 and 89 % respectively. The cost of coal is 25, 28 and 31 US$/ton
respectively. Determine the coal mix that will minimize the fuel cost meeting all the
77 Energy efficiency Optimization for Power plants. Faculty : Dr.G.G.Rajan
constraints , at a steam generation rate of 750 t/hr of steam. In this example, we shall
develop the Linear equations for the above mentioned constraints.
Formulation of the problem.
Let x1,x2 and x3 be the quantity of coal used in the process.
Heat output by x1 tons of coal / hr = x1 * 1000 * 4000 kcal/hr.
Heat absorbed by boiler = x1 * 1000 * 4000 * 0.86 kcal /hr.
Steam Generated by x1 t/hr of fuel = (x1 * 1000 * 4000 * 0.86) / (780*1000)
(Enthalpy of steam is 780 kcal/kg) = 4.410256 x1 t/hr
Steam Generated by x2 t/hr of fuel = (x2 * 4500 *0.87 ) /780 = 5.01923 x2 t/hr
Steam Generated by x3 t/hr of fuel =(x3 * 4900*0.89) / 780 = 5.591025 x3
The sum of these 3 equations should be equal to 750 t/hr.
Steam Generation constraint:
i.e. 4.410256 x1 + 5.01923 x2 + 5.591025 x3 >= 750 (I)
Sulfur Content Constraint.
(3.8 x1 + 1.9 x2 + 1.0 x3) / (x1+x2+x3) < = 1.8
i.e. 2.0 x1 + 0.1 x2 –0.8 x3 <= 0 (II)
Phosphorous Content constraint.
0.02 x1 + 0.01 x2 + 0.005 x3 / (x1 + x2 + x3) <= 0.015
i.e. 0.005 x1 – 0.005 x2 – 0.010 x3 <= 0 (III)
Ash content constraint:
3.0 x1 + 2.0 x2 + 5.0 x3 / (x1 + x2 + x3) <= 3.0
i.e. 0 x1 –1.0 x2 + 2.0 x3 <= 0 (IV)
Objective function :
Minimize 25 x1 + 28 x2 +31 x3
Let s1,s2,s3 and s4 be the slack variables added to equations I to IV. The final LP
problem is represented by
Minimize 25 x1 + 28 x2 +31 x3
ST
78 Energy efficiency Optimization for Power plants. Faculty : Dr.G.G.Rajan
4.410256 x1 + 5.01923 x2 + 5.591025 x3 + S1 = 750 (Ia)
2.0 x1 + 0.1 x2 –0.8 x3 –s2 = 0 (IIa)
0.005 x1 – 0.005 x2 – 0.010 x3 – s3 = 0 (IIIa)
–1.0 x2 + 2.0 x3 – s4 = 0 (IVa)
To get the initial feasible solution from the objective function, x1 should be maximized to
minimize the fuel cost.
LP program output
Min = 25 * x1 + 28 * x2 +31* x3;
4.410256* x1 + 5.01923 * x2 + 5.591025 * x3 >= 750;
2.0*x1 + 0.1* x2 -0.8* x3 < = 0;
0.005 * x1 - 0.005* x2 - 0.010* x3 <= 0;
-1.0 * x2 + 2.0 * x3 <=0;
Global optimal solution found at iteration: 3
Objective value: 4174.802
Variable Value Reduced Cost
x1 0.000000 0.4507412
x2 95.97245 0.000000
x3 47.98623 0.000000
Coal Wt tons Wt fraction
Scontent
Phosporous Ash % Steam gen t/hr
x1 0 0 x2 95.97245 0.66667 1.9 0.01 2.0 5.01923 x3 47.98623 0.33333 1.0 0.005 5.0 5.591025Total 143.95868 1.0000 1.6 0.0083 2.99999 750 Targets 1.80 0.015 3.000 750
Check for constraints fulfillment .
79 Energy efficiency Optimization for Power plants. Faculty : Dr.G.G.Rajan
For solving the above equations Ia through IVa and getting the optimal values,
readers are advised to refer to any standard Operations Research / Linear Programming
books listed in the reference. Above solution has been obtained by using Lingo 8
program, which may be downloaded from www.lindo.com. EORT Models have been
developed using SCIMOD software or Excess spread sheet programming.
Energy Efficiency Optimization Techniques - Overall System
Any complex system comprises many subsystems, some energy intensive, some
not so intensive and others least intensive. Energy efficiency optimization of the overall
system may be achieved by two methods.
By overall optimization of the total energy system data or by synthesizing the
individual sub system data into one consolidated data.
A typical example is the energy consumption of a complex refinery which
comprises many sections or sub processes which has certain process objective. Each unit
consumes energy resources such as fuel, steam and power.
Using the operating data, specific energy consumption of each unit may be
determined. This may be synthesized to arrive at overall energy consumption of the
process. In case of overall optimization method, all the energy cost centers will be taken
into the data as input parameters and a multivariable model developed to determine the
total energy consumption of the process. The observation is given in the table below. The
advantage with individual unit specific consumption is that it is possible to identify
problem areas / units more precisely for remedial action.
The overall energy consumption is the sum of individual unit consumptions. This
may vary with plant capacity utilization, equipment efficiency and other operating
parameters. In the example given below there are sixteen process units including steam
generation units in the refinery. Assuming 5 critical parameters per unit, total number of
variables to be used in the overall model shall be eighty. To arrive at a fairly accurate
80 Energy efficiency Optimization for Power plants. Faculty : Dr.G.G.Rajan
model, a minimum of 80 to 90 observations are required. This is a very time consuming
process and requires extensive computer programs to solve.
The approach to modeling and optimization application is similar to EORT modeling
discussed earlier. For successful overall system energy optimization, following steps are
to be followed.
UNIT T'PUT STEAM FUEL POWER EXPORT EXPORT SP.NRG Name t/hr t/hr t/h mw STM t/h PWR mw '000/t crude 500 28.5 9.55 1275 0 0 241.5
vb 100 12.8 5.65 1585 1.2 0 686.5 cru 60 35 2.05 1200 12.5 0 643.5
hds1 100 6.5 3.21 1185 0 0 391.4 hds2 120 8.5 3.55 1275 0 0 367.9 hds3 100 6.9 3.45 1285 0 0 420.2 crude 400 25 8.05 1495 0 0 257.0 vac1 100 12.8 5.65 1585 0 0 693.9 vac2 500 28.5 9.55 1275 0 0 241.5 bitum 100 12.8 5.65 1585 0 0 693.9 treatr1 50 8.5 1.55 1075 0 0 459.6 treatr2 75 10.8 1.65 1385 0 0 344.8
lpg 35 10.5 1.55 1175 0 0 697.9 blr1 100 12.8 6.65 1585 90 0 240.9 blr2 100 15.5 6.15 1275 85 6000 183.5 blr3 100 13.8 6.05 1385 85 0 214.0
Note: Total Energy Consumed kcal/h : 8.527796E+08 Equivalent Fuel in kg/h : 81217.11 Fuel Calorific Value in kcal/kg 10500 Steam Enthalpy for consumption 680 Steam Enthalpy for generation 620 Specific Energy is given as : '000 kcal/ton
81 Energy efficiency Optimization for Power plants. Faculty : Dr.G.G.Rajan
Data collection / validation / reconciliation / transfer to / modeling input files
Error detection and elimination
Parameter Estimation
Performance Monitoring - generation of Key Performance Indicators (KPI)
Optimization - improvement of efficiency and profitability
Modeling - representation of plant using mathematical models
This may be automated or manually done based on the complexity of the process and
requirement. When more number of parameters are to be used in the optimization model,
the data may be transferred from DIDC console using suitable interface modules. This
solves the problem of manual data entry. When the parameters are less, manual data entry
and processing may be done.
Since the objective of the whole exercise is energy efficiency optimization, only the
energy related parameters need to be involved in the model. To identify performance
deterioration of a particular unit, the specific energy consumption value given in the last
column may be compared to the standard / basic case.
If the deterioration is very critical, corrective actions may be taken to restore the
energy efficiency. A time dependant model may be used to indicate energy efficiency
trend of all energy intensive sub systems as explained earlier.
Energy efficiency optimization - benefits.
Optimization of equipment performance against “base case ‘ or ‘zero date ‘ or design
conditions allows the operator to assess the current and future operation against the
energy demands imposed by the site-wide balances. Proper evaluation of current
equipment performance is a vital component in the optimization strategy, providing a
more accurate energy efficiency of the equipment due to potential changes in operating
variables.
82 Energy efficiency Optimization for Power plants. Faculty : Dr.G.G.Rajan
M. Optimal Operation.
The primary concern of any utilities optimization system is to satisfy the site-wide
energy balances. In simple terms, all process unit and internal utilities plant requirements
must be achieved. Utilities optimization is not generally concerned with minimizing
process unit energy requirements; these are essentially fixed by operating conditions.
Process unit energy consumption should, however, be considered within any site-wide LP
or rigorous optimization strategy which defines the current operating conditions.
Optimization aims at meeting the requisite energy at the minimum cost within the
constraints imposed on the process by energy balances, environmental considerations and
utilities equipment limitations. These have been shown in a number of examples and case
studies.
Once these energy demands are satisfied, benefits are delivered by operating the
utilities process in the most cost-effective manner. This involves determining the
optimum loading for individual items of equipment.
Typical examples are
Power generation on each turbine set.
Boiler steam generation rates.
Secondary set-points such as turbine pass-out rates.
Allocation of fuel across the available fired boilers.
Given the constantly changing nature of the energy balances, this problem may be
typically solved by an automated system .
Optimal Selection
In addition to the benefits generated by optimizing the existing utilities
configuration, the operator can consider an alternative scenario whereby the operating
set-up is changed. For example:
Determine the optimum number of operating turbine sets and line-up.
83 Energy efficiency Optimization for Power plants. Faculty : Dr.G.G.Rajan
Choose between steam and electricity to drive auxiliary devices.
Make a decision on the best choice of fuel types.
This problem is typically solved less frequently and must consider equipment
availabilities together with any costs associated with altering the current configuration.
Optimization of Maintenance and Planning activities.
A properly applied process model will reflect current equipment performance.
This information may be used to compute the cost of degraded performance. A
comparison is made between the current degraded performance and the “day-zero” or
base case / clean performance. (e.g. performance of condensers / coolers etc)
The incremental cost associated with degraded performance is used together with
the anticipated costs of equipment cleaning .
These costs include:
Maintenance charges, which are generally well known.
Outage cost of the maintenance period, which is more complex and can be
computed by the optimizer.
Equipment maintenance strategies can be better planned if trends are developed and used
as a basis for evaluating future performance and costs against anticipated energy
requirement profiles. From these figures, the operator can determine the correct time for
equipment cleaning or replacement.
Another challenge facing the process industry is how to effectively manage
planned shut-down of equipment for statutory inspections or plant turnarounds.
Extending the optimization to include “what-if” scenario evaluation allows the user to
develop appropriate strategies to deal with the substantial changes in energy balances that
result from the shutting down of large-scale producers or consumers.
The Integrated Approach
Fig 2.07 shows this approach of energy efficiency optimization of the total
system. This was explained in chapter 1 with specific examples.
84 Energy efficiency Optimization for Power plants. Faculty : Dr.G.G.Rajan
Process Units Optimization. This level is concerned with individual process areas within the Refinery /
Petrochemical complex and is an activity carried out frequently , depending on the
process being considered. The primary activity is Performance Monitoring of the major
steam and electricity generators and users, on an individual basis. The individual units are
modeled in a sophisticated and rigorous manner. Each equipment model contains one or
more parameters with explicit engineering meanings that can be updated to reflect the
current performance of the units. This section may be considered the base level, on top of
which higher-level functionality is built. These parameters are used in the higher-level
Optimization calculations. Using these parameter values , it is possible to evaluate and
predict equipment performance / deterioration and plan for maintenance. Hence it is
possible to maintain the energy efficiency of the total system at it’s peak.
Fig 2.07. Optimizing total energy of the system
85 Energy efficiency Optimization for Power plants. Faculty : Dr.G.G.Rajan
A very important feature of optimization is the calculation and prediction of the
various restricted emissions, such as NOx, SOx and Cox which were covered in some of
the case studies. This may be evaluated for the total system. Simplified models, based on
regressed plant data relate process unit heat and power usage against production rate.
When this information is integrated for the entire units, the model output shall give the
total steam, power, water, air demand etc. This output shall be used in the utility
optimization model. These relationships are reduced to comparatively simple “specifics”
– equations used by the production planning department and expressed as “MW/ te of
product”, or “te of HP steam / te of product”.
Utilities Requirements
If an online system is used, optimization model coefficients may continuously be
upgraded to provide better level of accuracy required for optimization. These models
should be flexible to incorporate different operating modes, which can significantly alter
the quantities of energy required by a process unit. Indeed, certain process units can
switch between net export and net import depending on the operating mode.
Optimization calculations, operating around the updated process models,
determine the optimum operating conditions for the equipment currently online using the
fuels presently in use.The optimizer operates within a feasible region bounded by
operating constraints such as
Site-wide energy requirements
Equipment limits and environmental targets.
Simultaneous balancing of all steam, fuel and power demands is performed at this level,
with minimal operating cost objective for the optimizer. Operating cost typically includes
cost of different fuels, electricity import / export, cost of de-mineralized water. The
technique may be used for equipment selection / maintenance decisions. Here, the
optimum combination of all available units may be evaluated to meet the current site
86 Energy efficiency Optimization for Power plants. Faculty : Dr.G.G.Rajan
demands. The operating cost function may be extended to incorporate penalties for
starting up and shutting down equipment units. This is designed to establish a realistic
pattern of equipment selection that can be effectively implemented by the utilities
operators.
Power Balance
Optimization can be extended to consider the possibilities of power import and export.
Performance of electrical equipment, balancing the load etc may also be carried out by
optimization.
87 Energy efficiency Optimization for Power plants. Faculty : Dr.G.G.Rajan
Case Study – Total Energy Optimization
This explains the application of simple optimization exercise on a micro level
operation of a petrochemicals plant using a single variable throughput. (load factor). All
the process parameters such as the feed quality, pressures, temperatures etc are taken as
constant. A series of models are developed to determine the fuel, steam and power
demand including loss. Fig 2.08, 2.09 and 2.10 show the output of models in graphical
form for fuel, steam and power demand as a function of unit throughput.
Fig 2.07 shows the consumption of fuel in ‘000 tons per ‘000 tons feed to the unit.
These values represent the actual consumption of fuel based on the operating data of the
Fig 2.08. Fuel Consumption model for an aromatics unit
10.0
12.0
14.0
16.0
18.0
20.0
22.0
t'put 000 tons
fuel
'000
tons
non linear 11.80 12.00 15.33 18.23 19.79
polynomial 11.39 12.69 14.95 17.96 20.15
actual 11.36 12.75 14.79 18.24 20.00
598 675 870 960 1000
88 Energy efficiency Optimization for Power plants. Faculty : Dr.G.G.Rajan
unit. The calorific value of the fuel is taken as 10000 kcal /kg and is known as Standard
Refinery Fuel. This considers the effect of fuel mix and its calorific value as an imaginary
fuel with a net calorific value of 10000 kcal/kg. This is the standard practice in Refinery
Engineering evaluations.
Fig 2.09 shows the consumption of steam in ‘000 tons per ‘000 tons feed to the unit.
These values represent the actual consumption of steam based on the operating data of
150
170
190
210
230
250
270
t'put '000 t
stea
m '0
00 t
non linear polynomial actual
non linear 169.7 171.6 207.2 238.7 255.7
polynomial 169.2 172.5 206.7 238.3 256.2
actual 167.8 175.5 200.0 249.6 250.0
598 675 870 960 1000
Fig 2.09. Steam Consumption model for an aromatics unit
89 Energy efficiency Optimization for Power plants. Faculty : Dr.G.G.Rajan
the unit. Steam pressure to the unit was 15 kg/cm2 g at 320 oC. Condensate recovery
from the unit was 40 %. This information may be used to determine the specific energy
consumption of the unit at various throughput levels.
Fig 2.10 shows the electrical power consumption of the aromatics unit at various
throughput levels. This represents the actual data of an aromatics plant. This may be used
in the specific energy consumption model, using appropriate conversion factors.
80.0
90.0
100.0
110.0
120.0
130.0
140.0
150.0
160.0
t'put '000t
pow
er '0
00 m
w
nonlin polynomial actual
nonlin 93.4 105.6 133.7 145.3 150.2polynomial 92.0 107.9 132.4 144.4 151.4actual 91.7 108.5 130.9 147.0 150.0
598 675 870 960 1000
Fig 2.10. Power Consumption model for an aromatics unit
90 Energy efficiency Optimization for Power plants. Faculty : Dr.G.G.Rajan
During the operation of any process plant, consuming liquid and gaseous fuels,
some quantity of fuel will always be lost in the form of flare loss or vent loss. In
aromatics production plant, additional losses encountered are hydrogen and solvent used
for extraction.
There are two methods of evaluating the specific energy consumption of the
plant. They are given by
Specific fuel and loss method and
Specific fuel consumption method.
Difference between the two values is denoted by specific energy loss of the plant,
a potion of which is controllable and the balance uncontrollable.
For reducing this energy loss, certain technological / operational changes may be
required and must be determined by cost-benefit analysis of the system on a case to case
basis. Fig 2.11 shows the hydrocarbon loss from the aromatics unit example. For this
purpose, actual plant data from a running unit has been considered.
Advantages of ARU energy consumption models:
These models shown in the Aromatics Unit example are the simplest and does not require
any rigorous process calculations and only the minimum plant data will be used for
developing these. Since this approach considers only the feed rate and other energy input
parameters independently, the plant energy consumption may be obtained by
synthesizing these values by incorporating an additional conversion program.
From these models, it is easy to identify areas of higher energy consumption
energy wise or cost wise from the operating parameters. Area of deterioration may be
pinpointed for corrective action at the right time.
91 Energy efficiency Optimization for Power plants. Faculty : Dr.G.G.Rajan
A total energy consumption model for various units using energy synthesis approach is
given in the table in previous pages.
4.0
4.5
5.0
5.5
6.0
6.5
7.0
7.5
t'put '000 tons
hc lo
ss '0
00 to
ns
actual nl model poly model
actual 5.21 5.31 5.40 6.09 7.20
nl model 5.39 5.02 5.49 6.39 6.93
poly model 5.19 5.36 5.30 6.26 7.11
598 675 870 960 1000
Fig 2.11. Hydrocarbon Loss model for an aromatics unit
92 Energy efficiency Optimization for Power plants. Faculty : Dr.G.G.Rajan
Methodology for conversion of energy inputs into specific energy consumption is
shown in the table below.
470.0
480.0
490.0
500.0
510.0
520.0
530.0
540.0
550.0
t'put 000 tons
sp.e
nerg
y '0
00 k
cal/t
sp.energy
sp.energy 540.7 529.9 478.7 524.5 524.0
598 675 870 960 1000
Fig 2.12. Actual Specific energy consumption for an aromatics unit
Specific total energy consumption - Aromatics Plantt'put sp.fuel sp.steam sp.power sp.energy 000t kg/ton kg/ton kw/ton 000 kcal/t 598 19.0 280.6 153.3 540.7 675 18.9 260.0 160.7 529.9 870 17.0 229.9 150.5 478.7 960 19.0 260.0 153.1 524.5
1000 20.0 250.0 150.0 524.0
93 Energy efficiency Optimization for Power plants. Faculty : Dr.G.G.Rajan
It may be seen from fig 2.12, that specific energy consumption of the unit is
minimum at the throughput of 870000 tons capacity. Using the energy cost details, it is
possible to determine the optimum plant load that will minimize the total energy cost.
Probably, that may show a different scenario as the energy cost of each input is likely to
be different depending on the energy cost at the generation source. For example, if the
steam and power are supplied from waste heat cogeneration system, the cost shall be
lower, than if they were generated by conventional fuel fired cogeneration system.
Solvent Loss Model.
In the Aromatics production unit, a solvent such as NMP or Sulfoline is used to
extract the aromatics from the aromatics / non aromatics solution after the conversion
stage. These solvents are volatile and are recovered in a recovery column. Loss of solvent
is minimized by maintaining proper bottom temperature in a steam heated reboiler. Loss
of solvent is also a form of energy loss in an aromatics unit. This also tends to vary with
the capacity utilization of the plant. Fig 2.13 shows the solvent loss as a function of
throughput of the Aromatics plant under consideration. Fig 2.14 shows the specific
solvent loss actual for the aromatics plant under study.
From the data given in the label, it is possible to determine the specific solvent
loss for the aromatics plant. Once this is done, it becomes easier to determine the
optimum plant capacity at which the operating profit will be maximum.
94 Energy efficiency Optimization for Power plants. Faculty : Dr.G.G.Rajan
Using the above specific consumption details, the unit operating capacity may be
optimized as shown below.
2.50
3.00
3.50
4.00
4.50
5.00
5.50
t'put '000 t
solv
ent '
000
t
actual nl model poly model
actual 2.93 2.77 3.48 4.32 5.00
nl model 2.95 2.74 3.47 4.40 4.94
poly model 2.92 2.79 3.44 4.39 4.96
598 675 870 960 1000
Fig 2.13. Solvent Loss model for an aromatics unit
95 Energy efficiency Optimization for Power plants. Faculty : Dr.G.G.Rajan
3.80
4.00
4.20
4.40
4.60
4.80
5.00
5.20
t'put 000 tons
solv
ent l
oss k
g/to
n
actual
actual 4.90 4.10 4.00 4.50 5.00
598 675 870 960 1000
Fig 2.14. Specific Solvent Loss model for an aromatics unit
96 Energy efficiency Optimization for Power plants. Faculty : Dr.G.G.Rajan
Total operating cost model
From the specific consumption models for fuel, steam, power, hydrocarbon loss
and solvent loss it is possible to develop a costing model which may be used to develop a
profitability model for the unit which may be optimized using the techniques mentioned
earlier .
If Cf, Cs,Cp,Chc & Cs are the cost of fuel, Steam, power, hydrocarbon and
solvent per ton respectively and specific consumption functions are f(x),s(x),p(x),h(x)
and l(x) respectively, then the total costs incurred are
Ct = Cf*f(x)+Cs*s(x)+Cp*p(x)+Chc*h(x)+Cs*l(x)
This is called ‘cost function’ which may be optimized to minimize the total cost.
As could be seen from the cost function equation, total cost is determined by the
individual cost of each energy component and the consumption pattern. For example let
us take just three energy components given by the equation
Ct = Cf*(a1*x2+b1*x +c1 ) + Cs*(a2*x2+b2*x+c2)+Cp*(a3x2+b3*x+c3)
Differentiating the above equation with respect to x and equating to zero, we get
dct /dx = (2 *Cf *a1 * x + b1) + (2 *Cs *a2 * x + b2) + (2 *Cp *a3 * x + b3)
= 2 x* (Cf *a1 + Cs *a2 + Cp *a3) + (b1+b2+b3) = 0
i.e. x = -(b1+b2+b3) / 2* (Cf *a1 + Cs *a2 + Cp *a3)
Differentiating the above differential w.r.t ‘x’ again, and substituting the value of
x, if the second differential value is negative, then x is maximum. If the value is positive,
then the value of x is minimum.
If the value of x fulfills the constraints stipulated, then x represents a feasible
solution. Else another value of x nearest to the lowest value of x in the constraint will
become the optimal solution.
Table given below shows the specific consumption models for fuel, steam, power, hc loss
and solvent loss for the Aromatics plant example.
97 Energy efficiency Optimization for Power plants. Faculty : Dr.G.G.Rajan
Cost of fuel, steam, power , hydro carbon mixture and solvent are used to develop
a cost function model as shown below.
Sp cost = 0.15*(5.900398E-05 * (x)^2 - 9.345668E-02 * x + 55.790424) + 0.02 *
(7.923447E-04 * (x)^2 - 1.328423 * x + 813.914264)+ 0.3* (1.934422E-04 * (x)^2 -
.3310468 * x + 298.171512) + 0.3 * (3.645079E-05 * (x)^2 - 6.303739E-02 * x +
34.19038)+ 2.50 * (2.872752E-05 * (x)^2 - 4.560455E-02 * x + 22.01971).
On simplification this is reduced to
Sp.total cost = 0.000165484188* x^2 - 0.272823594 * x + 55.049275
Following table shows the workings of arriving at the final coefficients of the Cost
function model.
Specific Consumption Models - Aromatics Unit
spfmax = (5.900398E-05 * (x)^2 - 9.345668E-02 * x + 55.790424)
spstmmax = (7.923447E-04 * (x)^2 - 1.328423 * x + 813.914264)
Sppwrmax = (1.934422E-04 * (x)^2 - .3310468 * x + 298.171512)
Hclossmax = (3.645079E-05 * (x)^2 - 6.303739E-02 * x + 34.19038)
Sp.solvmax = (2.872752E-05 * (x)^2 - 4.560455E-02 * x + 22.01971)
Cost of fuel = 150 us$/t = 0.15 us$/kg
Cost of steam = 20 us$/t = 0.02 us$/kg
Cost of power = 300 us$/mw = 0.30 us$/kw
Cost of hc = 300 us$/t = 0.30 us$/kg
Cost of Solvent = 2500 us$/t = 2.50 us$/kg
98 Energy efficiency Optimization for Power plants. Faculty : Dr.G.G.Rajan
This on differentiated with respect to x and equating to zero gives
0.000330968 x = 0.272823594.
i.e. x = 0.272823594 / 0.000330968 = 824.3 (‘000 tons)
Second order differential gives a positive value. Hence the value of x is
minimum. This shows that at 824.3 (‘000) tons of throughput level, the total cost of
energy components will be minimum.
This is a deterministic model where the outcome is given by a cost function
equation. It may be noted, that in arriving at this cost function, all the cost parameters
have been given a firm value.
Since the cost of these energy components are dynamic, the optimum level of
operation will shift from time to time. Hence developing a cost function program will be
very useful for taking care of cost dynamics.
Fig 2.15 shows the specific total cost model derived from these observations.
cost fn x^2 fn x fn const
0.15 8.8506E-06 -0.0140185 8.3685636
0.02 1.58469E-05 -0.0265685 16.278285
0.30 5.80327E-05 -0.0993140 89.451454
0.30 1.09352E-05 -0.0189112 10.257114
2.50 7.18188E-05 -0.1140114 55.049275
total 0.000165484 -0.2728236 179.40469
99 Energy efficiency Optimization for Power plants. Faculty : Dr.G.G.Rajan
This example shows the importance of optimization in plant operation which will
minimize the total energy cost and maximize net operating profit of any process. Most
important part of the whole exercise is the data must be consistent and reliable. In many
cases, data reconciliation must be applied. However reliability of the model depends on
the number of observations used in the model.
67
68
69
70
71
72
73
t'put '000 MT
sp.to
tal c
ost u
s$
sp.total cost
sp.total cost 67.87 67.06 67.07 67.91 69.57 72.07
750 800 850 900 950 1000
Fig 2.15. Specific total cost vs throughput – Aromatics Unit.
100 Energy efficiency Optimization for Power plants. Faculty : Dr.G.G.Rajan
Impact of process variables:
Accuracy of the models may be improved , by incorporating additional process
parameters which have an impact on energy consumption of the process. In the case of
Aromatics Unit, feed quality represented by PONA analysis ( Paraffin, Olefin, Naphthene
and Aromatic content) plays an important role in energy consumption.
Hence the models may be modified into a two variable model of the form
E = a * (t’put)b * (N+A)c
where E is the energy consumption , N and A are volume % of naphthenes and aromatics
present in the feed. In the single variable model, all these parameters are assumed to be
constant and only the capacity utilization has been considered. Other parameter which
affects energy consumption in the Aromatics Unit is the hydrogen/hydro carbon ratio of
the recycle gas ( hydrogen purity), Severity of reforming operation and catalyst activity.
An EORT model covering these parameters may be developed to determine
energy consumption as a function of conversion.
Case study – optimum insulation thickness
This is a very common optimization problem encountered in insulation technology. In hot
product lines such as steam pipelines, hot oils etc thickness of insulation has to be
optimized. If x is the thickness of insulation, fixed increases with thickness. But cost of
heat loss reduces.
Hence the total cost which is the sum of fixed cost and variable cost reaches an
optimum level, at which the total cost must be minimum. Equations may be written for
these costs as given below.
Cf = a * x + b
Where Cf is the fixed cost, x is insulation thickness and a & b are constants.
Cost of heat loss may be written as
Chl = c/x + d
101 Energy efficiency Optimization for Power plants. Faculty : Dr.G.G.Rajan
Where Chl is the cost of heat loss, c and d are constants.
Total cost Ct = Cf + Chl = a * x + b + c/x + d
Differentiating this function w.r.t x we get
dCt / dx = a – c/x2 = 0
x2 = c/a
x = c/a
Differentiating the function again, we get d2Ct / dx2 = + 2*c/3*x3 = +ve
Hence x value represent the minimum point in the total cost function.
Optimization procedure – single variable.
The first step in optimization is to determine which parameter has to be
optimized. Typical factors could be total cost per unit of production (e.g. Aromatics Unit
case discussed above), profit, % conversion etc. Once the basis is determined, relations
between parameters as a function of objective variable must be developed by regression
methods or process models. Finally these relationships are combined to give the desired
optimum conditions. This has been shown in the insulation thickness. In this the objective
function is to determine optimum insulation thickness which will minimize the total cost .
The variables are the fixed costs, variable costs which are combined to give the total cost.
The problem was solved by analytical method using differential calculus.
Optimization procedure – two or more variables.
When two or more variables are involved in optimizing an objective function, the
procedure for optimization is tedious. However the procedure is similar to a single
variable problem. Out of the two independent variables, one variable is kept constant and
the other optimal variable is determined.
Let us consider the total cost function given by
Ct = ax + b/xy + cy + d
102 Energy efficiency Optimization for Power plants. Faculty : Dr.G.G.Rajan
Where a,b,c & d are positive constants.
The objective function is to find out the values of x and y which will minimize the total
cost given by the above equation. For this purpose, partial derivatives of x & y keeping
one variable constant is carried out as shown.
( Ct / x) = a – b/x2y = 0
( Ct / y) = -b/xy2 + c = 0
Solving these two equations, we get x = (cb/a2)1/3 and y = (ab/c2)1/3
When more than two variables are involved, the same procedure is adopted to determine
the value of independent variables.
Let us assume that the range of x variable is 10 to 20 and the range of y is 12 to 36 and
the coefficients in the cost function are 2.33, 11900,1.86 and 10. When these values are
substituted in the equation, the distribution of x, y and Ct will be as shown in the
following table.
Table. Optimizing two variables ( x and y) in the above example
This gives an optimum value of x = 16 and y =20. Total Cost is 121.7
y values
12 18 24 30 36 20
x
10 154.8 132.9 127.5 128.8 133.3 130.0
11 148.1 129.2 125.3 127.5 132.6 126.9
12 142.9 126.5 123.9 126.8 132.5 124.7
13 138.9 124.6 123.1 126.6 132.7 123.3
14 135.8 123.3 122.7 126.8 133.2 122.3
15 133.4 122.5 122.6 127.2 133.9 121.8
16 131.6 122.1 122.9 127.9 134.9 121.7
17 130.3 122.0 123.4 128.7 136.0 121.8
18 129.4 122.1 124.1 129.8 137.3 122.2
19 128.8 122.5 125.0 130.9 138.6 122.8
20 128.5 123.1 126.0 132.2 140.1 123.6
103 Energy efficiency Optimization for Power plants. Faculty : Dr.G.G.Rajan
Analytical method gives x = 15.97 and y = 20.01.
This problem may be solved by any NLP solvers like Lingo. The problem may be
formulated as
Minimize 2.33 * x + 11900 /x *y + 1.86 * y + 10
Subject to
x >= 10
x < 20
y >= 12
y < 20
Table given above shows the methodology adopted analytically, using the same
programming steps.
Practical Energy Efficiency constraints in optimization
This is an important factor that must be considered in energy efficiency
optimization problems. These constraints should be fixed with thorough knowledge of the
energy generating equipments like turbines, boilers, heaters etc. Impact of process
parameters on energy efficiency must be understood clearly.
In the case of gas turbine for example, the limitations are fuel input, heat release,
air/fuel ratio, power output at terminals and exhaust gas temperature. From pollution
point of view, Sulfur level in the fuel also forms a constraint. In the case of Heat
Recovery Steam Generator, the constraints are waste gas flow rate, its temperature, stack
temperature, steam generation rate etc. An EORT model may be developed using these
operating parameters and this will form the objective function. Since most of the energy
intensive equipments deteriorate with use, these constraints also will undergo a change
depending on the existing condition of the hardware configuration, which will ultimately
affect the optimal operating parameters. This is valid for almost all equipments in
operation and optimization is a dynamic exercise that must be conducted regularly to
minimize the overall energy cost.
104 Energy efficiency Optimization for Power plants. Faculty : Dr.G.G.Rajan
Impact of energy efficiency on production cost and profitability.
It was proved in the examples and case studies that energy efficiency optimization
does reduce the production cost and increases profitability. Tangible results could be
obtained only when optimization is carried out on class A equipment series, as they
consume mote than 70 to 75% of total energy input. Besides this, waste heat recovery
must be optimized wherever possible, as this amounts to energy recycling and the
specific energy consumption and energy cost will be minimum at optimum energy
recycling. This is a dynamic exercise and the optimum point is determined by the energy
cost equivalent of fuel.
Impact of energy efficiency on environment
Efficient heater / boiler operation conserves fuel and simultaneously reduce
pollution . However, for every fired heater / boiler an optimum excess air has to be
maintained, below which energy conservation may tilt the balance to increase pollutants (
in the form of Carbon monoxide , Soot, SPM etc) as given in the fuel mix example. When
the excess air is reduced, heater / boiler efficiency starts improving and SO2 emissions
will start coming down as shown in figs 2.16. After certain point, fuel input will start
increasing, as the entire fuel will face air deficiency and will result in incomplete
combustion Fig 2.16. SO2 emission in kg/hr vs Heater efficiency.
105 Energy efficiency Optimization for Power plants. Faculty : Dr.G.G.Rajan
The flue gas will indicate presence of Carbon monoxide, soot etc as shown in fig
2.17. In this typical case, the flue gas does not contain any CO or soot at 10% excess air.
When an attempt is made to reduce the excess air to 5,2.5 & 1 %, CO in the flue gas
starts increasing to 3000, 5000 & 12000 ppm respectively. This indicates incomplete
combustion of fuel. Under these conditions, smoke could be seen from the stack.
It is obvious from this example, that each heater has an optimum excess air, below
which pollution starts increasing in the form of soot, CO etc. Similarly at high excess air
rates also pollution will start increasing due to lower efficiency and higher fuel
consumption for the same heat duty. Hence efficient heater operation could be achieved
by operating the heaters / boilers at optimum excess air level. Optimum excess air is a
dynamic parameter that varies with the design, life, burner condition etc. These
parameters need to be monitored on a continuous basis to achieve high efficiency and low
Fig 2.17. CO emission in ppm vs % Excess air
106 Energy efficiency Optimization for Power plants. Faculty : Dr.G.G.Rajan
emissions. Carbon monoxide is a toxic gas and could endanger the environment if
adequate care is not taken. In the example given above , while excess air reduction from
10% to 1% reduces emission from 37.2 ton/h to 34.32, CO pollution starts increasing
from 0 ppm to 12000 ppm which is higher than the allowable limit of 1200 ppm. Above
is a clear case of sub optimal operation. These factors must be taken into consideration in
fixing the constraints while optimization exercise is carried out.
Optimal fuel mix by LP Model
Fuel mix plays an important role in controlling atmospheric emissions. Cleaner fuels cost
more than the dirty stocks, but reduce pollution control costs and offer longer equipment
life and better efficiency. An optimum mix may be arrived at using LP techniques and
NLP models as shown in the fuel mix examples. In general, impact of energy efficiency
on environment should be considered in the constraints.
Impact of energy efficiency on safety
Safety is of utmost importance in optimization exercise. There are many parameters,
which are likely to be unnoticed / sacrificed during energy efficiency optimization. It is a
good practice, if the constraint relationships include safety parameters quantitatively
which may be judged by past experience and historical accident data.
Practical constraints in equipment, sub systems and systems.
A number of constraints may be faced in real life situations during operation.
There are occasions in which the energy efficiency has to be sacrificed for meeting the
production targets, at a lower overall efficiency. A less efficient system of heaters /
boilers will have to be necessarily operated to meet the production target. It is the judicial
choice of the operations chief that will determine the implementation of optimal
parameters.
EORT models will be an answer to fixing the operational constraint limits, as his is built
on actual operating parameters under various conditions. When once the model is
developed, it is possible to determine the feasibility of meeting the constraints by a trial
107 Energy efficiency Optimization for Power plants. Faculty : Dr.G.G.Rajan
run at the stipulated parameters. Hence, plant parameter optimization exercise shall be
carried out successfully after getting the concurred values of the various constraints.
Another problem experienced in the complex plant is the down stream unit capacity /
equipment sizing, which may not match the upstream process flows / conditions leading
to a non optimal condition. This is a challenge to be tackled by the optimization group
which may warrant a rigorous process study.
Conclusions
In this chapter we have covered the basic principles of optimization that may be
used in real life situations. A number of examples such as boiler loading, heater
efficiency optimization, fuel mix optimization, EORT modeling etc were explained with
live cases. A typical Aromatics Unit example was shown in detail as to how the capacity
of the plant may be optimized for maximum operating profit. These methods may be
applied in real life situations to get the best out of competing resources.
108 Energy efficiency Optimization for Power plants. Faculty : Dr.G.G.Rajan
1. Rajan.G.G., "Computer-aided Utility Management', Paper, International
Conference on Computer Simulation, Oregon, May.1996
2. W.L.Nelson, 'Petroleum Refinery Engineering ', McGraw-Hill Book Company,
Inc. New York
3. David M.Himmelblau, 'Process Analysis by Statistical Methods', John Wiley &
Sons, Inc. New York
4. Dr.Rajan.G.G., 'Optimal Management of Resources in Process Industries ', Indian
Society for Technical Education, April 1982
5. Technical Audit - User's Mnual ', Techno Software International, India.
6. Arnold H.Boas, 'Modern Mathematical Tools for Optimisation', Reprint Chemical
Engineering, McGraw-Hill Book Company, Inc. New York
7. Dr.Rajan.G.G., 'Energy Efficiency Improvement - Decision Models for selection
of alternates and cost effective solutions', International Conference on Energy,
WEC, New Delhi. 12th Jan '96.
8. Dr.Rajan.G.G.,'Operational Analysis of Process Industries for Productivity
Improvement', Productivity, NPC,New Delhi, July-Sep '89
9. Dr.Rajan.G.G., 'Computer-aided Process Decisions', Summer Computer
Simulation Conference, Ottawa,Canada, May 1995
10. Dr.Rajan.G.G., 'ERP Models for the Hydro Carbon Processing Industry', MRL-
CHT Oil Industry Computer Meet, Chennai, April '99..
11. Samuel E.Bodily, 'Modern Decision Making', McGraw-Hill Book Company, Inc.
New York
12. Max S.Peters, Klaus D.Timmerhaus, Plant Design and Economics for Chemical
Engineers, McGraw Hill Kogakusha Ltd, Toranto
13. Gupta P.K , Hira D.S., Operations Research , S.Chand and Company, New Delhi,
India
14. Loomba M.P., Linear Programming , Tata McGraw Hill Publishing Company
Ltd, New Delhi, India.
15. Rajan.G.G., 'Computer-applications and Decision Support System for Efficient
Management of Industries',PhD. Thesis, BITS,Pilani, 1996
16. Owen L.Davies,Peter L.Goldsmith, ' Statistical Methods in Research and
Production', Longman,New York
References
Energy Efficiency Optimization for power Plants Training Program – Dr.G.G.Rajan
109
POWER PLANT MANAGEMENT
3
Energy Efficiency Optimization for power Plants Training Program – Dr.G.G.Rajan
110
3. POWER PLANT MANAGEMENT
Introduction :
A modern power plant has a typical layout as shown in fig 3.01. Power plant design has
a number of options .The design criteria , based on process economics could be
A coal fired boiler ( pulverized coal firing type, Fluidized bed combustion
type )
An oil fired boiler or
A gas fired boiler
A waste heat boiler.
Thermal Power Plant
In a thermal power plant, steam is produced and used to spin a turbine that operates a
generator. Shown here is a diagram of a conventional thermal power plant, which uses coal, oil,
or natural gas as fuel to boil water to produce the steam. The electricity generated at the plant is
sent to consumers through high-voltage power lines.
Energy Efficiency Optimization for power Plants Training Program – Dr.G.G.Rajan
111
In thermal power plant operation, each equipment contributes to the overall energy
efficiency of the plant. Since efficiency of each equipment contributes to the overall efficiency of
the power plant, the total system has to be optimized for achieving the maximum efficiency at
required power demand, which is dynamic.
Boiler system :
Factors which affect the performance of boiler may be denoted by an energy supply chain
concept. Starting from the fuel, till steam generation a number of parameters influence boiler
efficiency.
Energy Resources
All Energy Consuming Systems / Devices / equipment consume Energy in one of the
following forms or a combination depending on the technology and / or requirement.
They are
Fuel Energy
Electrical energy
Steam energy
Thermal energy
Total boiler system may be split into sub systems such as fuel system, Air system, BFW
system, boiler main , steam system, Flue gas system and heat recovery system. Each subsystem
performance has an impact on overall boiler efficiency. Parameters that affect the performance
of each subsystem is given in the following table.
Energy Efficiency Optimization for power Plants Training Program – Dr.G.G.Rajan
112
Processparameter
Associatedattribute
Parameters Key indicators
Fuel Solid
Liquid
Gas
Type, c/h ratio, particle size, ash content, burner efficiency., temperature
Density, c/h ratio, viscosity, ash content, atomization efficiency., temperature
Gas composition , burner efficiency, temperature
Good quality and particle size improve combustion efficiency.Combustion zone temperature, air / fuel ratio, specific fuel consumption under identical conditions.
,, ,,
,, ,,
Air Air purity, humidity and temperature, excess air quantity
Air contamination and humidity reduces efficiency . High air temperature increases efficiency. Pre heated air saves fuel burnt in the boiler for the same load.
Key indicators : Combustion zone temperature, excess air for identical process conditions. Reduces FD fan load and energy consumption.
Burner system Burner efficiency
Fuel temperature / viscosity, Atomizing steam to fuel ratio, Fuel pressure, nozzle diameter
High fuel temperature and low viscosity improves burner efficiency. Steam to fuel ratio must be optimum. Flame length, color and absence of soot are visible parameters. Combustion zone temperature is an indication of good combustion.
Combustion system
Combustion efficiency
Flame temperature and good flame pattern
Proper combustion increases boiler efficiency all other parameters remaining constant.
Energy Efficiency Optimization for power Plants Training Program – Dr.G.G.Rajan
113
Boiler Main BFW quality
TDS should be within allowable limits. pH of BFW should be between 7 to 7.2.
If TDS becomes excessive, water tubes will be blocked with Ca / Mg salts and may lead to tube failure. If pH is very low tube failure will be caused by corrosion.
Economizer Economizer efficiency. Higher the temperature pick up of temperature by BFW, higher will be the efficiency of boiler. and lower the flue gas temperature
Any fouling on economizer tubes, will retard heat transfer and increase fuel consumption in boiler. Economizer surface must be as clean as possible.
Energy Efficiency Optimization for power Plants Training Program – Dr.G.G.Rajan
114
BOILER THEORY AND OPERATING PRINCIPLES
Steam generator is a pressure vessel into which liquid water is pumped at the operating
pressure . After the heat has vaporized the liquid , resulting steam is then ready for use or for
further heating in a super heater . A typical section of a water tube boiler ( end view from drum
side ) is shown in fig 3.01 .Hot boiler feed water (bfw) enters the boiler drum at a pressure
slightly higher than the drum pressure .The water gets circulated in the water tubes which are
heated by the hot gases passing through the combustion chamber .
During circulation, water is converted into steam and flashes in the upper portion of the
drum.This steam will be at it's saturated temperature corresponding to the operating pressure
of the drum .This steam passes through a set of coils into the combustion chamber, where it gets
super heated to the required temperature by contact with hot gases produced in the combustion
chamber due to combustion of fuel .
Super heater is a steam header which is placed in the combustion chamber as shown in
the figure . After getting super heated , the steam may be used in turbine for generating
power / work or may be used in process as per the heat requirement . For maintaining the
required super heated steam temperature , a de- super heater is placed outside the boiler
Energy Efficiency Optimization for power Plants Training Program – Dr.G.G.Rajan
115
chamber , where boiler feed water is added at a lower temperature and required quantity , by
a flow control device.
During boiler feed water circulation from the drum to lower pot and vaporization , the
concentration of hardness producing solids increase and tend to produce severe scales in the
tubes , drum walls etc . This tendency retards heat transfer and thermal efficiency of the
boiler . These scales reduce flow rates in the coil section and cause tube failure also. To over
come this problem , a part of the water is blown down periodically and in some cases
continuously from the drum , based on the boiler feed water hardness and quality. This is a
source of energy loss , as excess of ' blow down ' corresponds to excess fuel input to the boiler.
From energy management point of view, thermal efficiency of the boiler which is
defined as the ratio of steam generated to the thermal energy input may be evaluated by
direct method or in-direct method as in the case of a heater. Fig 3.02 shows the energy input
and output streams for a boiler, used for thermal efficiency calculation.
Energy Input to boiler ( fuel side) :
Enthalpy + Enthalpy + Heat ofin fuel in air combustion
Energy losses ( fuel side ) :
Enthalpy in + heatflue gases losses ( convection & radiation )
Energy Efficiency Optimization for power Plants Training Program – Dr.G.G.Rajan
116
Thermal Efficiency of a boiler is determined by one of the following methods.
Method A : from heat balance
Method B : from heat losses
In either case , the thermal efficiency may be expressed in terms of either net or
gross heating value of the fuel . Heat output in the case of a boiler is the heat value of steam
generated by the boiler less the heat value of feed water and steam returned to the unit.
Heat input includes the heat value of fuel used by the unit ( based on net or gross
calorific value) plus any other waste heat that may be supplied to the boiler from an external
source.
Energy Efficiency of boilers
A heat balance of boiler indicates the type of heat losses which could be used for
identifying areas of improvement . A number of parameters affect the thermal efficiency of
the boiler as in the case of a heater. They are
* Sensible heat in flue gases
* Convection and Radiation losses.
* Unburnt Combustibles in flue gases.
* Unburnt combustibles in refuse
* Blow Down losses etc
3..2.1.Sensible heat in flue gases:
This loss is the largest in a boiler ( refer box 4.10 ) and represents the heat carried
away by the flue gases and released to the atmosphere without doing any useful work .
Hence it is obvious that if more than the required quantity of air is used in a boiler ,
more will be the loss in flue gases and the thermal efficiency of the boiler will be reduced
correspondingly.
Energy Efficiency Optimization for power Plants Training Program – Dr.G.G.Rajan
117
3.2.2 Convection and Radiation losses :
This depends on the temperature of boiler's external surface . Quantity of heat lost by
convection and radiation is a function of shell temperature and wind velocity. This loss occurs
basically due to poor insulation and poor design characteristics . If the refractory lining and
other insulating materials are not in good condition , these losses will tend to increase and
reduce the thermal efficiency of the boiler. Hence it is imperative that the surface temperature
at various sections should be monitored periodically to minimise this loss for timely action.
3.2.3 Unburnt matter in flue gas:
When combustion is incomplete , part of the carbon present in the hydro carbon fuel
may be converted into carbon monoxide instead of carbon-di-oxide.When the carbon content
of the fuel is not fully burnt to carbon-di-oxide, there will be substantial energy loss to the
atmosphere from the flue gases ( besides atmospheric pollution) , the loss being proportional to
the amount of carbon monoxide produced .This could be estimated by a carbon balance across
the boiler taking into consideration carbon content in fuel , carbon-di-oxide and carbon
monoxide in the flue gases and carbon content in the refuse.
Unburnt matter may show itself in the flue gas in the form of black smoke which
represents presence of carbon particles in the flue gas. It may also show itself as Carbon
monoxide in the flue gas . Heat loss due to incomplete combustion may be estimated from the
heat of combustion data as given below.
C + O2 = CO2 + 8084 kcal / kg of carbon
2C + O2 = 2CO + 2430 kcal / kg of carbon
While every kg of carbon present in the flue gas represents a loss of 8084 kcals , every kg of
Carbon partially oxidized to CO results in a loss of 5654 kcals . This situation could be
controlled with optimum excess air input to the boiler.
Energy Efficiency Optimization for power Plants Training Program – Dr.G.G.Rajan
118
Another source of loss is from the refuse formed which is applicable to solid fuels .
This is estimated by evaluation of refuse formed per unit weight of fuel burnt and from the
analysis of combustibles in the refuse .
Unburnt matter for normal efficiency calculation is taken as Carbon. For example if
Wr is the quantity of refuse produced per unit weight of fuel and if C is the weight fraction of
carbon in refuse, then heat lost in the refuse is given by
Hr = Wr x C x 8084 kcal / kg fuel burnt
3.2.4 Blow Down Losses
Dissolved salts find entry to the boiler through make-up water which is continuously fed
by the Boiler Feed Water pump ( bfw) . In the boiler , there is continuous evaporation of water
into steam . \
This leaves behind the salts in the boiler. Concentration of these salts , tend to increase
in the boiler drum and starts precipitation after certain concentration level .
Water from the drum should be blown down to prevent concentration of salts beyond
certain limits . Since the water in the boiler drum is at a high temperature ( equivalent to it's
saturation temperature at boiler drum pressure ) , excess blow-down will lead to loss of energy
known as 'blow-down losses ' .
Blow-down rate reduces the boiler efficiency considerably as could be seen from
fig 3.-03.
Hence it is imperative that blow-down rates are optimized ,based on the hardness levels
of boiler drum water which is a function of the operating pressure.
Energy Efficiency Optimization for power Plants Training Program – Dr.G.G.Rajan
119
Example 1 – Power Generation
In power generation a number of combinations of energy resource may be
deployed.
Conventional Power Generation Route is given below.
Factors that affect power generation efficiency are
Combustion Efficiency
Boiler Efficiency / Steam transmission efficiency
Turbine Efficiency and
Generator efficiency.
In this case optimization refers to fuel, Steam and Power system
Mathematical Model
Mathematically, the overall efficiency of the micro power Generation system may be
represented by
E overall = E comb * E boiler * E turbine * E generator
In a typical case following
conditions were observed as
against design values .
Observed Design
Combustion efficiency 1.00 1.00
Boiler Efficiency 0.85 0.91
Turbine Efficiency 0.40 0.45
Generator Efficiency 0.95 0.975
Overall System efficiency 0.3230 0.39926
Productivity as % of base 0.8089 1.0000
Energy Efficiency Optimization for power Plants Training Program – Dr.G.G.Rajan
120
Efficiency Deviations
In the example
Actual boiler efficiency is lower by 16%
Turbine efficiency is lower by 5%
Generator Efficiency is lower by 2.5 %
Overall system efficiency is lower by 6.626 %
Productivity is lower by 19. 11 %
This aspect is seldom noticed by energy users.
In power plant operation , this will be observed in terms of excessive fuel consumption
and high operating cost.
Power plant management :
An effective power plant management system covers efficiency optimization of
o Fuel system
o Boiler feed water system
o Combustion control system
o Boiler furnace
o Air pre heater / Flue gas control system
o Super heater system
o Economizer system etc.
Following formats show the amount of information required for evaluating the boiler
efficiency by indirect method.
When boiler efficiency is calculated periodically, it is possible to identify efficiency
deterioration of a particular boiler in a power plant for corrective action.
Power plant operating economics depends of the fuel cost, steam / fuel ratio and power
generation quantity. Hence turbine efficiency also has to be monitored periodically. A rule of
thumb is the specific steam consumption / mw power generated.
Energy Efficiency Optimization for power Plants Training Program – Dr.G.G.Rajan
121
1. Boiler Efficiency evaluation by indirect method
Design case A User : ABC Power plant B Unit : TT Power Division C Equipment : Boiler no 6
Date of Observation Stream Day : Run Date : Run No from start
D I.Fuel Data (coal) Weight Liqfuel
fraction Carbon Hydrogen Moisture Oxygen Sulfur Nitrogen Ash
E II.Observed Process parameters
Boiler Duty mmkcals/h % Unburnt matter in refuse Amb.Temperature oC Flue Gas Temp o oC Relative Humidity %
F II a. Flue gas data
O2 vol % CO2 vol % CO vol % SO2 vol % Excess Air
Energy Efficiency Optimization for power Plants Training Program – Dr.G.G.Rajan
122
Fuel High Heating value Kcal/Kgm Fuel Low Heating value Kcal/kgm
G III.Energy Loss data
e.Radiation Loss : f.Blow Down Loss : g.Unaccounted Losses h.Loss due to combustibles
Total Loss % (on dry basis)
H IV.Boiler Efficiency design a.EFFICIENCY HHV basis b.EFFICIENCY LHV basis It is preferable that observed data is provided for various load conditions.
II. Economizer performance evaluation design Observed data and date
1 Flue gas flow t/hr * to be calculated
2 inlet temperature oC 553 3 outlet temperature 362 4 BFW flow t/hr 290 5 BFW inlet temp oC 230 6 BFW outlet temp oC 292 7 Area of heat transfer m2
III. Air preheater performance evaluation design Observed data and date
1 Flue gas flow t/hr
2Flue gas inlet temperature oC 362
3 outlet temperature 147 4 Total Air flow t/hr 313.75 5 Primary air flow t/hr 128.6 6 Secondary air flow t/hr 185.15
Energy Efficiency Optimization for power Plants Training Program – Dr.G.G.Rajan
123
7 Air inlet temp oC 30 8 Air outlet temp oC (PRI) 303 9 Air outlet temp oC (sec) 317
10 Area of heat transfer m2
IV. Superheater performance evaluation design Observed data and date
1 Flue gas flow t/hr * to be calculated
2 SH inlet temp oC (stage I) 1140 3 SH stage II inlet temp oC 997 4 SH outlet oC 743 5 Steam ( SH outlet ) t/hr 290 6 Steam inlet temp oC 7 SH steam outlet temp oC 8 Area of heat transfer m2
Energy Efficiency Optimization for power Plants Training Program – Dr.G.G.Rajan
124
V. Steam turbine performance evaluation design Observed data and date
1 Steam inlet to turbine t/hr 288.61
2 I stage Steam flow t/hr 18.91 ,, pressure kg/cm2 31.76 ,, temp oC 372.6
2 II stage Steam flow t/hr 17.47 ,, pressure kg/cm2 17.46 ,, temp oC 296.1
3 III stage Steam flow t/hr 19.92 ,, pressure kg/cm2 7.781 ,, temp oC 257
4 IV stage Steam flow t/hr 19.09 ,, pressure kg/cm2 2.199 ,, temp oC 123.3
5 V stage Steam flow t/hr 11.18 ,, pressure kg/cm2 0.511 ,, temp oC 81.9
6 Final stage Steam flow t/hr 201.96 ,, pressure kg/cm2 0.107 ,, temp oC 47.2
7 Power Generated in MW 74
8 Heat rate KJoule/KWH 9359.5 Kcal / KWH 2235.5 Efficiency % 38.47 Weighted TD efficiency %
Energy Efficiency Optimization for power Plants Training Program – Dr.G.G.Rajan
125
Typical boiler efficiency calculation
Date : 12.10.07 Stream Day : 200
Input Data
item unit quantity
Fuel Data
1.Fuel t/hr 5.0000
2.Type Liquid
3.Density g/ml 0.9976
4.Gross CV kcal/kg 10800
5.Fuel temp oC 98.5
Flue Gas
1.CO2 % 12.7
2.O2 % 3.65
3.CO % 0
4.N2 % 83.65
5.Temperature oC 215
6.Air temp oC 30
Atomising Steam
1.Steam Flow t/hr 0.75
2.Pressure kg/cm2 10
3.Temperature oC 260
Boiler Feed Water
1.BFW Flow t/hr 78.5
2.Pressure kg/cm2 110
3.Temperature oC 118
Steam Generation
1.Steam Flow t/hr 74.5
Energy Efficiency Optimization for power Plants Training Program – Dr.G.G.Rajan
126
2.Pressure kg/cm2 100
3.Temperature oC 320
Blow Down
1.Blow Down t/hr 3
2.Pressure kg/cm2 100
3.Temperature oC 320
Design Duty
1.Heat Duty mmkcal/h 60
Material Balance
input output
1.Fuel t/h 5.0000
2.A.Steam t/h 0.7500
3.Dry Air t/h 81.7923
4.moisture t/h 0.0009
Total t/h 87.5432
Flue Gases
CO2 t/h 16.3443
CO t/h 0.0000
O2 t/h 3.7276
N2 t/h 62.5171
SO2 t/h 0.1000
H2O t/h 4.4332
error term t/h -0.4211
Total t/h 86.7011
Energy Efficiency Optimization for power Plants Training Program – Dr.G.G.Rajan
127
ENERGY BALANCE
Basis: 1 kg fuel/hr
input
1.Heat of combustion kcal/hr 10800.0
2.Heat in pre heat kcal/hr 36.5
3.Pre Heat in Air kcal/hr 58.9
4.Heat in atomising
steam kcal/hr 103.8
Total Heat input kcal/hr 10999.2
output
1.Heat Gain in BFW kcal/hr 1617.1
2.Steam Generation kcal/hr 8344.9
3.Dry Gas Loss kcal/hr 802.0
4.Water formed kcal/hr 483.1
5.Air moisture kcal/hr 1.6
6.Blow Down kcal/hr 179.1
7.Incomplete Combn kcal/hr 0.0
8.Others kcal/hr 0.0
9.Error term kcal/hr -428.6
Total Heat output kcal/hr 10999.2
Reconciled output
%
1.Heat Gain in BFW kcal/hr 1556.4 14.15
2.Steam Generation kcal/hr 8031.9 73.02
3.Dry Gas Loss kcal/hr 771.9 7.02
4.Water formed kcal/hr 465.0 4.23
5.Air moisture kcal/hr 1.5 0.01
6.Blow Down kcal/hr 172.4 1.57
7.Incomplete Combn kcal/hr 0.0 0.00
8.Others kcal/hr 0.0 0.00
Total Heat output kcal/hr 10999.2 100.00
Energy Efficiency Optimization for power Plants Training Program – Dr.G.G.Rajan
128
Boiler efficiency by direct
method = (8344.9+1617.1)/10999.2) x 100
= 90.57%
Energy Efficiency Optimization for power Plants Training Program – Dr.G.G.Rajan
129
Power Generation System
TURBINES
All process and power industries require power and / or steam for driving various types
of equipments which are design specific. These include steam turbines, gas turbines,
electric motors, hydraulic turbines, turbo-expanders, and engines. However steam and gas
turbines are the prime movers for most applications.
Turbines may be classified as in Table 3.1.
Classification of turbines
Table 3.1 Classification of turbines
Name Type Application a.Steam urbine
back-Pressuretype
power generation
Extraction cum condensing
power and steam
totalcondensingturbine
Power generation or as prime mover for compressors, pumps etc.
b.Gas turbine
For power generation and / or as prime mover
Steam Turbines
Dependability, variable speed-operation and possibility of energy savings are the
basic factors that are considered for choice of steam turbines for many process
applications . This is also determined by Heat-Power requirement of the process.
The steam turbine is a very satisfactory and dependable prime-mover for many
process machines such as pumps ,fans , blowers and compressors . It is often used as a
driver for electric generator to provide main power to the process plant motor drives .
Steam turbine's feature of variable speed is useful for saving energy on pump , blower,
compressor drives etc.
Energy Efficiency Optimization for power Plants Training Program – Dr.G.G.Rajan
130
Where process load fluctuations are frequently encountered, steam turbines offer
the best solution for energy saving. Steam turbines are more dependable and no process
interruptions, power failures, transmission problems etc are ever faced in case of
steam turbines. It is a common practice to keep steam driven equipment running in
critical services, where power interruptions may cause a serious problem .
Types of steam turbines
All conventional process - plant steam turbines are axial-flow turbo machines in
which steam flows parallel to the shaft axis. These turbines may be of single stage or
multi stages. In single stage turbines, steam expands through a set of nozzles only once.
They are most suited for smaller applications where a few horse-power of power (say 50
to 150 HP) is required. On the other hand, multi stage turbines are used where the
requirement of power is very high.(say 1000 to 2000 hp ). Multi stage turbines have
two or more expansions through a set of nozzles in each stage.
When the exhaust steam from any steam turbine is above atmospheric pressure,
the turbine is called non-condensing. When the exhaust steam is below atmospheric
pressure, this is labeled condensing turbine. Using second law of thermodynamics and
Bernoulli's Equation, it may be seen that
V 12 v 2 2
------- + h1 = ------- + h2 + W 2g J 2g J
where
v 1 and v 2 are velocity at inlet and outlet in ft /sec respectively
h 1 and h 2 are enthalpy at inlet and outlet conditions of steam in
btu / lb respectively
g = gravitational constant ( 32.2 ft/s2)
J = Joule's constant ( 778 ft-lb / btu)
The velocity of steam flowing at inlet and outlet conditions is approximately
equal. Hence v 1 may be taken as equal to v 2 in the above equation.This reduces the
above equation to
Energy Efficiency Optimization for power Plants Training Program – Dr.G.G.Rajan
131
h 1 - h 2 = W
i.e. Work done by the steam turbine is directly proportional to the enthalpy difference at
inlet and outlet conditions. It may be noted that inlet and outlet steam conditions are
fixed. Hence enthalpy values may be found from the steam - tables or Mollier diagram
for the inlet and outlet conditions. However, it may be found that in actual practice
isentropic expansion is never achieved due to energy losses in friction and inefficiencies
of the turbine. The actual enthalpy will be slightly higher than the constant entropy value
but in the same pressure line.
Hence overall turbine efficiency per stage may be defined as
E stage = ( h 1 - h 2' ) / (h 1 - h 2 )
Where
h 2' = actual enthalpy observed at the outlet conditions.
A typical single stage back pressure steam turbine is given in fig 3.4 .
Performance of steam turbine is represented by Theoretical Steam Rates (TSR)
and Actual Steam Rates (ASR) which is the heat quantity in BTUs or KCALs required
to generate one kwh of power.
HP STEAM INLET
MP / LP STEAM OUTLET
~
Energy Efficiency Optimization for power Plants Training Program – Dr.G.G.Rajan
132
fig 3.4 . Back Pressure Turbine
Since 1 kwh = 3413 btu / hr or
= 860 kcal / hr
TSR = 3413 / ( h 1 - h 2 ) lb / kwh
ASR = 3413 / (h 1 - h 2 ' ) lb / kwh
Where
h 1 = enthalpy at inlet of the turbine in Btu / lb
h 2 = enthalpy at outlet of the turbine in Btu / lb (isentropic)
h 2 ' = enthalpy at outlet of the turbine in Btu / lb (actual)
In metric units, the same is represented by
TSR = 860 / ( h 1 - h 2 ) kg / kwh
ASR = 860 / (h 1 - h 2 ' ) kg / kwh
E stage = TSR / ASR
Above relationship could be monitored from the power generation and steam input,
output data. As can be seen from the above efficiency relationship, greater the steam
pressure drop through the turbine, greater will be the power output. A reduction in
exhaust pressure causes a greater power generation than an increase in the inlet steam
pressure. It may be noted that specific steam consumption depends on the absolute
pressure ratio of the turbine.
Back Pressure turbines are thermally efficient having cycle thermal efficiencies in the
range of 75 to 85 % (ref fig 3.4)
Condensing turbines tend to be high in cost and thermally inefficient - giving around 10
to 25 % efficiency.
This could be economically exploited, only in processes where large amounts of waste
heat is available and where no other recovery options are found lucrative. It is found
more energy efficient to exploit this waste heat from low pressure steam sources such as
evaporators to drive the local devices, than to generate power.
Energy Efficiency Optimization for power Plants Training Program – Dr.G.G.Rajan
133
Fig 3.5 is a typical condensing turbine . In a total condensing turbine, all the
steam entering the turbine is condensed in a surface condenser, which is located at the
exhaust section of the turbine.
Last stage of the turbine (surface condenser) is under vacuum and the
condensation takes place only in the surface condenser. Steam should not be allowed to
condense in the impeller blades, as this may cause rotor imbalances, Dew point corrosion,
and silica deposits in the rotor.
fig 3.5 total condensing turbine
Extraction cum condensing turbines is the most common type for total generation
schemes. This offers excellent flexibility to meet the steam and power demand of the
process and is energy efficient also. Cycle thermal efficiencies of this type is around 50 to
70 %.
~
HP STEAM
Surface
Condenser
Condensate
Energy Efficiency Optimization for power Plants Training Program – Dr.G.G.Rajan
134
Fig 3.6 shows an Extraction cum condensing turbine . This turbine combines
the function of both back pressure and condensing turbine. Part of the steam after doing
work is taken out at medium pressure for process use, while the balance quantity goes to
the surface condenser for condensation. Provision exists to vary the extraction steam to
condensing steam, by a pass out valve. Power generated or work done varies with the
extraction steam and condensed steam.
fig 3.6. Extraction cum condensing turbine
~
HP STEAM
EXTRACTIONSTEAM - MP
SURFACE
CONDENSER
Energy Efficiency Optimization for power Plants Training Program – Dr.G.G.Rajan
135
Gas Turbines:
Gas turbines are self-contained power plants. A gas turbine as the name implies uses gas
as the fuel , which burns with compressed air in the combustor and is expanded through the
turbine which drives the air compressor and delivers the excess of power generated as output
. In the last step, the products of combustion are exhausted to the atmosphere.
As soon as the fuel burns in the combustion chamber, the temperature of the hot gases reaches
a maximum. If t 1 is the temperature achieved by combustion and to is the ambient
temperature and q is the energy input to the turbine, then the maximum recoverable work is
given by
W max = (( t 1 - t 0) / t 1 ) * q
W loss = ( t 0 / t 1 ) * q
Eff = (t 1 - t o) / t 1
fig 3.7. Simple Cycle Gas Turbine.
Above relationship is based on 'Carnot Cycle ' which gives the highest possible
conversion rate of heat into work. Actual efficiency will be much lower than the Carnot
Efficiency for the same temperature and flow conditions. Thermal Efficiency of a gas
HOT GAS T1 oC
TURBINE
EXHAUST To C
WORK
Energy Efficiency Optimization for power Plants Training Program – Dr.G.G.Rajan
136
turbine depends on pressure ratio and turbine inlet temperature. Fig 7.4 shows the
typical arrangement of a simple cycle gas turbine.
Turbine efficiency
Turbine efficiency is the ratio of energy output to energy input irrespective of whether it
is steam turbine or gas turbine. This is calculated by an energy balance across the turbine. A
typical enthalpy balance for a multi stage extraction cum condensing steam turbine is similar to
fig 7.3, but the number of extraction may be two to three at different extraction pressures.
Material balance
w 1 = e 1 + e 2 + c 1 + l 1
where
w 1 = steam input to turbine in t/hr
e 1 = extraction rate at stage 1 in t/hr
e 2 = extraction rate at stage 2 in t/hr
c 1 = condensed steam in t/hr.
l 1 = steam losses in t/hr.
Energy Balance
Input = steam extraction I extraction II condensing loss input - steam - steam - steam - steam energy energy energy energy energy
Energy content of each stream is calculated from the enthalpy values at it's actual pressure and
temperature conditions. Hence the above relationship may be given by
E i = w1 * h 1 - ( w 2 * (h 1 - h 2) + w 3 * (h 1 - h 3) + c1 * (h 1 - h 4) )
where
h 1 = steam enthalpy at pressure p1 and temperature t1
h 2 = steam enthalpy at pressure p2 and temperature t2h 3 = steam enthalpy at pressure p3 and temperature t3
Energy Efficiency Optimization for power Plants Training Program – Dr.G.G.Rajan
137
h 4 = steam enthalpy at pressure p4 and temperature t4respectively
Output
E o = 860 * p
where
p is the power output in kw/h
860 conversion factor
E o is kcal/hr.
Turbine Efficiency (Actual)
Eff = ( E o / E i ) * 100
As a thermal prime mover, the efficiency of a turbine is the useful work energy that
appears as shaft power and is presented as a percentage of the chargeable heat energy .
The over-all thermal efficiency of a steam turbine is given by W / JQ where W is the
shaft work in ft lb or kg metre and Q is the btu or kcal chargeable. This leads to several
expressions for thermal efficiency as follows.
Where a number of extraction rates are carried out in a steam turbine , the efficiency is
given by
2545 Eff ' = ----------------------------------------------------------------------------------- w x h1 - ( w ' a x h a + w ' b x h b + w 'c x h c + ....) - w 2 x h f 2
where
w = inlet steam in t/hr
w 'a = Extraction steam stage I in kg/hr
w 'b = Extraction steam stage II in kg/hr
w 'c = Extraction steam stage III in kg/hr
w2 = Exhaust steam in kg/hr
h1 = Enthalpy of steam at inlet conditions kcal/kg
ha = Enthalpy at stage I extraction conditions kcal/kg
Energy Efficiency Optimization for power Plants Training Program – Dr.G.G.Rajan
138
hb = Enthalpy at stage II extraction conditions kcal/kg
hc = Enthalpy at stage III extraction conditions kcal/kg
hf2 = Enthalpy at exhaust conditions kcal/kg
Gas Turbines
Energy distribution in simple GT
A look at simple gas turbine process shows that most of the added fuel energy disappears
with the exhaust gas as shown below
Fuel Energy 100 %
Energy distribution pattern in a simple Gas Turbine
Only about 29% of the fuel energy is converted into useful power, since the exhaust gases
still have high temperature. For minimizing the energy losses, it is imperative to use the fuel with
the highest possible efficiency. This calls for utilization of the heat in the exhaust turbine gases
in the most economical and feasible way.
Combustion Chamber
TurbineElectrical
Electrical and mechanical
Exhaust Gas Losses
Energy Efficiency Optimization for power Plants Training Program – Dr.G.G.Rajan
139
Turbine Exhaust Temperature:
This may be calculated using the gas law and is given by
T2/T1 = ( P2/P1) ( -1)/ )
Where
o P2 & P1 are turbine exhaust and inlet pressures in absolute units
o T2 & T1 are exhaust and inlet temperatures in absolute units.
o is the adiabatic exponent.
e.g:
P1=4.5
P2=0.5(kg/cm2g)
T1 = 850 oC
y = 1.25
i.e T2/(850+273) = (1.523/5.523)^(1.25-1)/1.25): T2 = 594.9 oC
Carnot Cycle Method
When a gas expands adiabatically from a Pressure P1 to P2 and at inlet temperature of T1 oC it
may be noted that
1.Turbine exhaust temperature is a function of pressure ratio
2.For the same inlet temperature, turbine efficiency is a function of exhaust temperature.
3.As inlet temperature increases, turbine efficiency also increases for the same exhaust
temperature. This concept is shown in the next slide.
4.For the 4 inlet temperatures (i.e. 800,900,1000,1100 oC), efficiency is highest at 1100 oC,
followed by 1000 oC and so on for the same exhaust temperature. High inlet temperature and
pressure demands special metallurgy , which is cost intensive.
Energy Efficiency Optimization for power Plants Training Program – Dr.G.G.Rajan
140
THERMO DYNAMIC EFFICIENCY OF TURBINES ( Carnot Cycle )
10
15
20
25
30
35
40
45
50
650 600 550 500 450
exhaust oC
ther
mo
dyna
mic
eff
icie
ncy
800 oC 900 oC 1000 oC 1100 oC
Turbine Efficiency is the ratio of actual work output of the turbine to the net
input energy supplied in the form of fuel.
For stand alone Gas Turbines, without any heat recovery system the
efficiency will be as low as 35 to 40%.
This is attributed to the blade efficiency of the rotor, leakage through clearance
spaces, friction, irreversible turbulence etc.
Energy Efficiency Optimization for power Plants Training Program – Dr.G.G.Rajan
141
Increasing Overall efficiency:
Since Exhaust gases from the Gas Turbine is high, it is possible to recover
energy from the hot gas by a Heat Recovery Steam Generator.(HRSG in short) and use the steam for process.Various options are given in the next few slides.
Turbine Efficiency is a function of
* Running Hours
* Load factor
* Fuel type
* Combustor efficiency
* System Energy Loss etc.
It is possible to predict turbine efficiency using the models.
Part load operating efficiencies of gas turbine power plant is given in the next slide for various
hard ware configurations. Refer the book for more information
Energy Efficiency Optimization for power Plants Training Program – Dr.G.G.Rajan
142
Cogeneration
Cogeneration (also called as combined heat and power, CHP) is the use of a heat engine
or a power station to simultaneously generate both electricity and useful heat.
Conventional power plants emit the heat created as a by-product of electricity generation
into the environment through cooling towers, flue gas, or by other means. CHP or a bottoming
cycle captures the by-product heat for domestic or industrial heating purposes, either very close
to the plant, or —especially in Scandinavia and eastern Europe—for distribution through pipes to
heat local housing. This is also called decentralized energy.[1]
In the United States, Con Edison produces 30 billion pounds of steam each year through
its seven cogeneration plants (which boil water to 1,000°F/538°C before pumping it to 100,000
buildings in Manhattan—the biggest commercial steam system in the world.
By-product heat at moderate temperatures (212-356°F/100-180°C) can also be used in
absorption chillers for cooling. A plant producing electricity, heat and cold is sometimes called
tri generation or more generally: poly generation plant.
Cogeneration is a thermodynamically efficient use of fuel. In separate production of
electricity some energy must be rejected as waste heat, but in cogeneration this thermal energy is
put to good use.
Energy Efficiency Optimization for power Plants Training Program – Dr.G.G.Rajan
143
Masnedø CHP power station in Denmark. This station burns straw as fuel. The adjacent
greenhouses are heated by district heating from the plant.
Thermal power plants (including those that use fissile elements or burn coal, petroleum,
or natural gas), and heat engines in general, do not convert all of their available energy into
electricity. In most heat engines, a bit more than half is wasted as excess heat . By capturing the
excess heat, CHP uses heat that would be wasted in a conventional power plant, potentially
reaching an efficiency of up to 89%, compared with 55% for the best conventional plants. This
means that less fuel needs to be consumed to produce the same amount of useful energy. Also,
less pollution is produced for a given economic benefit.
Some tri-cycle plants have utilized a combined cycle in which several thermodynamic
cycles produced electricity, and then a heating system was used as a condenser of the power
plant's bottoming cycle. For example, the RU-25 MHD generator in Moscow heated a boiler
for a conventional steam power plant, whose condensate was then used for space heat. A more
modern system might use a gas turbine powered by natural gas, whose exhaust powers a steam
plant, whose condensate provides heat. Tri-cycle plants can have thermal efficiencies above
80%.
An exact match between the heat and electricity needs rarely exists. A CHP plant can
either meet the need for heat (heat driven operation) or be run as a power plant with some use of
its waste heat.
CHP is most efficient when the heat can be used on site or very close to it. Overall
efficiency is reduced when the heat must be transported over longer distances. This requires
heavily insulated pipes, which are expensive and inefficient; whereas electricity can be
transmitted along a comparatively simple wire, and over much longer distances for the same
energy loss.
Energy Efficiency Optimization for power Plants Training Program – Dr.G.G.Rajan
144
A car engine becomes a CHP plant in winter, when the reject heat is useful for warming
the interior of the vehicle. This example illustrates the point that deployment of CHP depends on
heat uses in the vicinity of the heat engine.
Cogeneration plants are commonly found in district heating systems of big towns,
hospitals, prisons, oil refineries, paper mills, wastewater treatment plants, thermal enhanced oil
recovery wells and industrial plants with large heating needs.
Thermally enhanced oil recovery (TEOR) plants often produce a substantial amount of
excess electricity. After generating electricity, these plants pump leftover steam into heavy oil
wells so that the oil will flow more easily, increasing production. TEOR cogeneration plants in
Kern County, California produce so much electricity that it cannot all be used locally and is
transmitted to Los Angeles[citation needed].
Types of plants
Topping cycle plants primarily produce electricity from a steam turbine. The exhausted
steam is then condensed, and the low temperature heat released from this condensation is utilized
for e.g. district heating.
Bottoming cycle plants produce high temperature heat for industrial processes, then a
waste heat recovery boiler feeds an electrical plant. Bottoming cycle plants are only used when
the industrial process requires very high temperatures, such as furnaces for glass and metal
manufacturing, so they are less common.
Large cogeneration systems provide heating water and power for an industrial site or an
entire town. Common CHP plant types are:
o Gas turbine CHP plants using the waste heat in the flue gas of gas turbines
o Combined cycle power plants adapted for CHP
o Steam turbine CHP plants that use the heating system as the steam condenser for the
steam turbine.
o Molten-carbonate fuel cells have a hot exhaust, very suitable for heating.
o Smaller cogeneration units may use a reciprocating engine or Stirling engine. The heat is
removed from the exhaust and the radiator. These systems are popular in small sizes
because small gas and diesel engines are less expensive than small gas- or oil-fired
steam-electric plants.
Energy Efficiency Optimization for power Plants Training Program – Dr.G.G.Rajan
145
o Some cogeneration plants are fired by biomass [5], or industrial and municipal waste (see
incineration).
Combined Heat Power Cycle / cogeneration:
Combined heat and power refers to the industrial units which generate their own
electricity and heat in the form of steam for meeting their heat and power demand of the
process
In most of the CHP systems, total heat and power is generated at an efficiency of
over 75 %, compared to the power generation efficiency of mere 30 to 35 %.
The optimum output of CHP is based on the total energy concept, in which the
primary fuel is used to generate both electricity and heat in one place.
Base case
Energy Efficiency Optimization for power Plants Training Program – Dr.G.G.Rajan
146
Energy Efficiency Optimization for power Plants Training Program – Dr.G.G.Rajan
147
Fuel Savings – CHP
Energy Efficiency Optimization for power Plants Training Program – Dr.G.G.Rajan
148
INTEGRATING TURBINE EXHAUST TO FURNACES
FUEL
AIRTURBINE
HEATER
BASIS: 1 TON FUEL AIR TH : 16.5 AIR ACT 28.0 EXH.TEMP 550 PRE-HEAT AVAILABILITY29*0.28*1000*(550-30)=4.22 MILLION KCAL/HR
Energy Efficiency Optimization for power Plants Training Program – Dr.G.G.Rajan
149
Energy Efficiency Optimization for power Plants Training Program – Dr.G.G.Rajan
150
Energy Efficiency Optimization for power Plants Training Program – Dr.G.G.Rajan
151
Waste Heat Recovery
o Refers to
o Recovery of Heat from Flue Gases
o Steam Condensates
o Waste Steam Recovery / Utilization
o Hot Waste Water & Process Streams
Waste Heat is the Heat that is rejected from a process.
Sources of Waste Heat are classified according to temperature in three ranges.
* High temperature range above 650 oC
* Medium temperature range between 250 to 650 oC
* Low temperature heat i.e. below 250 oC
Parameters to be evaluated for Waste Heat Recovery
* Temperature of Waste Heat Fluid
* Flow rate of waste heat fluid
* Chemical Composition of waste heat fluid
* Minimum allowable temperature of Waste Heat fluid
* Temperature / Chemical composition of cold fluid
* Maximum allowable temperature of Cold fluid.
* Control temperature if required for cold stream.
Energy Efficiency Optimization for power Plants Training Program – Dr.G.G.Rajan
152
Examples :
Waste Heat Recovery from CO Boiler
Energy Efficiency Optimization for power Plants Training Program – Dr.G.G.Rajan
153
Energy Efficiency Optimization for power Plants Training Program – Dr.G.G.Rajan
154
Air Pre Heater Lay Out for Waste Heat Recovery
Air-Pre Heating - Advantages
Air Pre-heating reduces fuel consumption by 25 %
Typical Case :
* Fired Heater Efficiency before APH installation: 70 - 71 %
* ,, ,, ,, after APH installation 85 - 89 %
* ,, ,, ,, with high Efficiency APH 91 %
* Reduction of 20 - 23 oC of Flue Gas temperature results in 1.% of fuel saving.
* Throughput can be increased by 20 %
Energy Efficiency Optimization for power Plants Training Program – Dr.G.G.Rajan
155
Energy Efficiency Optimization for power Plants Training Program – Dr.G.G.Rajan
156
TOTAL POWER PLANT OPTIMIZATION
Total power plant optimization involves optimization of
o BFW system including economizer
o Air preheating system
o Fuel / burner system
o Combustion control
o Boiler furnace control
o Steam super heating system
o Flue gas system
o Steam distribution system
o Turbine system
o Generator system and
o Transmission loss control.
Energy Efficiency Optimization for power Plants Training Program – Dr.G.G.Rajan
157
BFW Quality:
BFW quality plays an important role in boiler efficiency maintenance.
Dissolved salts find entry to the boiler through make-up water which is
continuously fed by the Boiler Feed Water pump ( bfw) . In the boiler , there is
continuous evaporation of water into steam . This leaves behind the salts in the
boiler. Concentration of these salts , tend to increase in the boiler drum and
starts precipitation after certain concentration level .
Water from the drum should be blown down to prevent concentration of salts
beyond certain limits . Since the water in the boiler drum is at a high temperature
( equivalent to it's saturation temperature at boiler drum pressure ) , excess blow-
down will lead to loss of energy known as 'blow-down losses ' . Blow-down rate
reduces the boiler efficiency considerably as could be seen from fig 3.8. Hence it
is imperative that blow-down rates are optimised ,based on the hardness levels of
boiler drum water which is a function of the operating pressure.
Energy Efficiency Optimization for power Plants Training Program – Dr.G.G.Rajan
158
In boiler operation practice , rate of blow down increases with steam pressure as the
scaling tendency increases with high temperature because the hardness limits are very
stringent .While fig 3.8 gives an estimate of % blow down on losses ,the same may be
calculated from the hardness levels of make-up water , flow rate ,steam generation rate and
the hardness level of drum water ( observed) .
Polynomial model given below could be used to determine the maximum limits of
TDS (total dissolved solids) that could be tolerated in the boiler drum operating at various
pressures .This is based on American Boiler Manufacturers' Association code of practice.
However, if the limits stipulated by the Boiler Designer is less than this value , the
lower of the two must be taken as the tolerance limit.
Fig 3.8
Energy Efficiency Optimization for power Plants Training Program – Dr.G.G.Rajan
159
Energy transmission system :
Total power plant management involves monitoring the performance of each sub system
by appropriate methods which is covered in the next few chapters.
Energy Efficiency Optimization for power Plants Training Program – Dr.G.G.Rajan
160
Calculating Energy Efficiency of Steam Turbines.
Steam turbines are devices which convert the energy stored in steam into rotational
mechanical energy. These machines are widely used for the generation of electricity in a number
of different cycles, such as:
Rankine cycle
Reheat cycle
Regenerative cycle
Combined cycle
The steam turbine may consists of several stages. Each stage can be described by
analyzing the expansion of steam from a higher pressure to a lower pressure. The steam may be
wet, dry saturated or superheated.
.
Consider the steam turbine shown in the cycle above. The output power of the turbine at
steady flow condition is:
Energy Efficiency Optimization for power Plants Training Program – Dr.G.G.Rajan
161
P = m (h1-h2)
where m is the mass flow of the steam through the turbine and h1 and h2 are specific enthalpy
of the steam at inlet respective outlet of the turbine.
The efficiency of the steam turbines are often described by the isentropic efficiency for
expansion process.
The presence of water droplets in the steam will reduce the efficiency of the turbine and
cause physical erosion of the blades. Therefore the dryness fraction of the steam at the outlet of
the turbine should not be less than 0.9.
Energy Efficiency Optimization for power Plants Training Program – Dr.G.G.Rajan
162
Isentropic Efficiency
The isentropic efficiency for an expansion process is defined as:
Energy Efficiency Optimization for power Plants Training Program – Dr.G.G.Rajan
163
Typical data for a condensing turbine
Energy Efficiency Optimization for power Plants Training Program – Dr.G.G.Rajan
164
EFFECT OF STEAM PRESSURE – CONDENSING TURBINE
From the figure given above, it may be noted that steam pressure has an impact on
specific steam consumption of the turbine. In this particular case, increasing steam
pressure from 35 to 45 kg/cm2, reduces steam consumption by 2.1%.
Hence steam boiler pressure is an important parameter in improving the energy efficiency
of the turbine.
Above information may be converted into turbine efficiency vs steam pressure,
taking the base case efficiency as given .
Energy Efficiency Optimization for power Plants Training Program – Dr.G.G.Rajan
165
EFFECT OF STEAM PRESSURE ON EFFICIENCY – CONDENSING TURBINE
As inlet steam pressure increases, turbine efficiency also increases. For an increase of
steam pressure from 35 to 45 kg cm2, turbine efficiency increases from 33.1 to 34.2 %.
Steam inlet temperature is also one of the parameters that determines the efficiency of
condensing steam turbine. Next figure explains this phenomena .
Energy Efficiency Optimization for power Plants Training Program – Dr.G.G.Rajan
166
STEAM INLET TEMPERATURE vs CONSUMPTION – CONDENSING TURBINE
As the steam inlet temperature increases from 300 to 400 oC, steam consumption %
drops from 100 to 88%.
In other words, steam turbine efficiency increases substantially as given in the next
diagram. Steam turbine efficiency increases from 33.1 to 34.4 % when inlet temperature
increases from 300 to 400 oC.
In Power plant management, these parameters may be optimized to get the best efficiency
possible, within the operational constraints..
Energy Efficiency Optimization for power Plants Training Program – Dr.G.G.Rajan
167
Exercise :
Following tables show the impact turbine operating parameters on efficiency. What are
the optimal conditions that can give the best efficiency of the turbine. What are the tools that
may be used in the evaluation process ? Maximum operating pressure of boiler 50 kg/cm2.
No Steam pressureKg/cm2
Turbineefficiency
in % 1 35 33.10
2 37 33.33
3 39 33.55
4 41 33.78
5 43 34.01
6 45 34.20
7 47 34.45
8 49 34.90
No Steam temperature
oC
Turbineefficiency
in % 1 300 33.10 2 320 33.35 3 340 33.55 4 360 33.80 5 380 34.10
6 400 34.35 7 420 34.61 8 440 34.85
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
160
4
PERFORMANCE MONITORING TECHNIQUES
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
161
4. Performance Monitoring Techniques
1. Monitoring Boiler Efficiency / Performance :
Boiler efficiency is determined by operation and maintenance practices of the total
system. Efficiency is also an indicator of mechanical condition of the boiler system as a whole.
Hence if the boiler efficiency is regularly monitored, it is possible to identify deterioration trend
and take corrective action. Using the deterioration trend, it is possible to predict the boiler
efficiency at a future period, by using historical models.
For boiler efficiency monitoring purposes either direct of indirect method or the
average may be selected and consistently used. A typical output of the boiler efficiency
program was given in the earlier chapter.
Precautions :Certain precautions are to be followed when boiler efficiency is determined as
listed below.
o Boiler load must be consistent.
o All flow meters must be calibrated and errors minimized.
o Pressure , temperature and draft profiles in the boiler circuit must be kept consistent
during the run.
While fig 4.01 gives a typical boiler efficiency curve as a function of load , impact of
running hrs on the efficiency is also given in the figure for various load conditions . Box 4.02
gives the performance models at base case and after ten months of operating time which may
be used for evaluation of performance and timely maintenance action.
Differentiating the above functions w.r.t load % and equating to 0 it may be found
that the maximum achievabe efficiency in the base case is 89.80 at 103.3% of design load ,
whereas the maximum achievable efficiency at the end of ten months of operation is 84.00 % at
a load of 99.16% .This clearly indicates , there is a drop in the efficiency of boiler with
the passage of time and load% .
This tends to affect the operating cost of the boiler and it is possible to determine
the optimum cycle length at which the boiler has to be shut down and cleaned so that the
efficiency may be restored to the level of base case .
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
162
Fig 4.01. Boiler Efficiency vs Load %
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
163
Performance monitoring technique:
Efficiency of boilers must be calculated for each boiler periodically and recorded
stream day wise. All the input parameters must be as accurate as possible and may be used as an
effective management tool to determine
o What is the current efficiency of the boiler ?
o What should be the efficiency at the current stream day and load %
o Whether the efficiency is within tolerable limits ?
o If not what are the reasons for lower efficiency ?
o What are the actual losses against base case ?
o Should the boiler be operated in the existing condition or should it be repaired ?
o Whether the boiler can be maintained or replaced etc.
A typical decision flow diagram to arrive at the right decision is given in fig 4.05 .
This methodology could be arrived at for any number of boilers when once the models are
developed for each controllable parameter .This uses a number of models for arriving at the
right decision.
Fig 4.05. Boiler Efficiency
management
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
164
Typical boiler efficiency program output
=================================================================
RESULTS - BOILER EFFICIENCY
=================================================================
User : ABC Refinery
Unit : India
Equipment : blr1
Input file : c:\tthermw\blr1.in
Output files: c:\tthermw\blr1.out
Output file2: c:\tthermw\blr1.mv
Archive file: c:\tthermw\boiler.arc
Date of Obs : 12.10.07
Stream Day : 100
Run Date : Run No : 1
I.Fuel Data
Carbon .87000
Hydrogen .12000
Moisture .00000
Oxygen .00000
Sulfur .01000
Nitrogen .00000
Ash .00000
II.Observed Process parameters
Fired Duty mmkcal/h 65.0000
% Heat load 100.0000
% Unburnt matter in refuse 0.0000
Amb.Temperature oC 30.0000
Flue Gas Temp oC 180.0000
Relative Humidity 0.0000
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
165
Excess Air 15.6454
Fuel High Heating Value - Kcal/Kgm 11163.8896
Fuel Low Heating Value - Kcal/kgm 10515.8887
III.Energy Losses
a.Dry Gas Loss : 5.2683
b.Air Moisture loss : 0.0000
c.Combustion Moisture Loss : 6.4719
d.Fuel Moisture Loss : 0.0000
e.Radiation Loss : 1.0000
f.Blow Down Loss : 2.0000
g.Unaccounted Loss : 0.3100
g.Loss due to combustibles : 0.0000
Total Loss % ( dry basis) 15.0502
IV.Boiler Efficiency
a.EFFICIENCY HHV basis : 84.9498
b.EFFICIENCY LHV basis : 90.1845
NOTE:
a.Air misture loss is due to moisture present in combustion air.
b.Combustion moisture is due to combustion of H2 in fuel to water.
c.Fuel moisture is due to presence of water in the fuel fired.
d.Radiation loss is due to heat loss from the exposed boiler surface.
e.Blow down is energy equivalent as % of fuel calorifiv value.
*** end of boiler efficiency program (indirect method) ***
Output information may be used to analyze performance of each boiler for efficiency
improvement.
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
166
A typical energy loss break-up for the boiler is given in fig 4.06. From this it is
obvious that all losses except wet losses show an increasing trend with respect to time.
Controllable losses in this case are dry gas loss , convection and radiation losses and blow
down loss .
Fig 4.06. Energy loss break up from boiler efficiency program output.
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
167
Observations:
o Increase in dry gas loss indicates gradual increase in excess air which could be due to
air leaks in the shell section . This is normally diagnosed by oxygen analysis at
combustion , pre-convection and post convection zones.
o Increasing trend in convection and radiation losses is indicated by high skin
temperature of the shell. This situation occurs due to thinning of the insulation brick. It
is a good practice to monitor this loss known as setting loss periodically, for deciding
the optimal replacement time.
o High blow-down rate could be due to poor operation practice or poor boiler feed
water quality . This may be checked from the TDS levels of make-up water , drum
water and allowable TDS limits.
Systematic recording and review of boiler parameters is very imperative to determine the
energy efficiency of a boiler.
A typical analysis data for a set of boilers is given in box 4.03. This is actually the
program output of boiler efficiency monitoring system program using a popular program,
which utilizes indirect method of determining boiler efficiency.
Blow down rate as % of energy input must be added to the last column to determine the
efficiency of boilers listed below.
Boiler Efficiency Models
Since boiler efficiency is determined by a number of parameters , it is imperative to
identify the impact of all parameters by suitable scientific methods. Boiler efficiency is
invariably affected by running days and load factor.
Scaling increases with the passage of time on both the water side and tube side. This
is reflected by rise in water tube skin temperature, flue gas temperature etc .
Since the boiler load is always dynamic depending on steam demand, the efficiency
could be estimated as a function of operating period and load % involving two variables .
Even a linear multi variable model is good enough for monitoring purpose. For this
purpose, common programs like MS Excel may be used .
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
168
Performance Monitoring of boilers
Performance monitoring of boilers is very imperative in power and process industries
as the major chunk of cost is governed by the fuel consumed in the boilers .
Since fuels in boilers are not in their pure form , they tend to deposit soot , scales etc in
the economizer , super heater , water tubes etc which retard heat transfer efficiency with the
passage of time .
In addition to this , steady loss in evaporating capacity is also experienced due to
accumulation of scales on the water side with the passage of time . The decrease in heat-
transfer rate is a function of time and is given by
where
U is the heat transfer coefficient in kcal /hr / m2 /oC
t is the time lapsed since last cleaning.
For minimizing the scale deposition on the convective section of the boiler, soot blowers
are used periodically to remove the loose deposits .However, hard scales tend to deposit with the
passage of time .
Hence performance monitoring of the boiler is essential to decide at what point of time
the boiler has to be taken out of service for cleaning and other maintenance jobs.
Like the case of heaters, boiler efficiency also varies with the passage of time due to the
above factors. In addition to this , boiler capacity utilization also affects the efficiency and
boiler manufacturers provide the boiler operating characteristics data.
Hence the performance of the boiler may be evaluated by the observed efficiency and
compared to the basic characteristics for the same operating conditions. Any abnormal
deviations observed could therefore be investigated to pin-point the problem area for
necessary action.
It is obvious from the above the boiler efficiency can be managed by monitoring the
efficiency of economizer, air pre heater and super heater by appropriate techniques.
1 / U2 = A + B * t
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
169
Specific Fuel Consumption
Specific fuel consumption is normally assumed to be a performance Indicator of energy
efficiency of the system. The system could be a heater or boiler or turbine. In conventional
Energy Management, Specific Energy Consumption is defined as the quantity of fuel consumed
in kgs or Lbs as given below.
Equipment Specific Fuel Consumption
1.Heater Kg fuel / Ton of feed processed
2.Boiler ton fuel / Ton steam produced.
3.Gas Turbine ton fuel / MW power generated.
4.Steam Turbine ton fuel eqvt steam / MW of Power
generated.
Examples given below are indicative of the concept of specific fuel consumption.
Specific fuel consumption may be used as a thumb rule to determine the equipment performance
/ efficiency by a quick glance. When all the parameters remain constant, specific fuel
consumption is definitely a simple and good performance indicator of the system under
consideration.
Item Feed tons Fuel tons Sp.Fuel
Consumption Unit
1.Heater 500 7.5 15 Kg/ton of
feed
2.Boiler 150 10 0.06667 Ton/ton of
steam prodn
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
170
3.Gas Turbine 25 MW 4.50 0.180 Tons of fuel /
MW
4.Steam
Turbine
10 MW 100 tons
steam
10 Tons steam /
mw
Historical data of specific energy consumption of system / systems in an indicator of
efficiency. With the passage of time efficiency starts dropping down, thereby increasing the
specific consumption of fuels.
Factors Affecting Specific Fuel Consumprion :
Fuel is consumed only in process heaters, boilers and Gas Turbines in the refining
industry. A number of process parameters affect specific fuel consumption in the case of heater /
boiler. They are
Capacity utilization and
Heater operating parameters such as
Draft
Transfer Temperature
Fuel type and it's Calorific Value
Fuel Mix ( % Oil , % Gas ) and
Combustion Efficiency of Fuel
Impact of capacity utilization on a fired heater efficiency was shown in earlier chapter when
all the other parameters are constant. Combustion Efficiency is a function of fuel mix used in a
heater / boiler.
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
171
Performance Monitoring of turbines.
Performance and condition monitoring of turbines are two different areas of energy
efficiency management. While condition monitoring refers to the mechanical condition of the
rotating equipment and is concerned with vibration level, bearing conditions , mechanical
condition etc. performance monitoring refers to the energy efficiency of the turbine under
consideration.
But it is necessary to know that condition of the rotating equipment has direct impact on
energy consumption and efficiency of the machine.The physical layout of various turbines were
presented in earlier chapters.
Performance Monitoring of steam turbines
While turbine efficiency is calculated based on the input/output principles, performance
of turbines vary with the load factor and aging / on -stream hours / days. Turbine
manufacturers often supply the performance characteristics of each machine in the form of
a curve for various energy input conditions.
Actual performance of the turbine may be compared with the base case to determine
whether or not the efficiency is within the acceptable limits. A typical steam turbine
characteristic is given in fig 4.5(a).
This characteristic curve gives the input steam rate vs output in kw for any chosen
extraction steam rate.
The graph is developed from the actual performance data for a small machine. For
performance monitoring of the turbine , the output observed at any load is compared against
the base case ( in this case characteristics curve) and checked whether the performance is
normal or not.
Case given in fig 4.5(a) has extraction rates of 0 ,5 ,10 and 15 t/hr and any intermediate
value is interpolated.
An example of how the turbine characteristics may be used for performance monitoring
is given in the section.
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
172
fig 4.5(a) .Typical Steam turbine characteristics.
observed data: steam inlet 20.8 t/h
pass - out 10.0 t/h
power output 2100 kw
From the characteristics, power output should be 2750 kw.
Observed Deviation in performance is -650 kw
% deviation from base value = (-650/2750)* 100
= 23.60%
The deviation is very high and needs a thorough investigation. If the turbine operation
continues, excess steam consumption would have been about 3.0 t/hr for the same output. (
given by y-x )
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
173
Increase in annual steam consumption based on 8000 hrs of operation works out to
24000 tons. For applying this monitoring technique, user must have the complete turbine
characteristic for each machine and refer to the same every time the checking is required.
Computer-aided performance monitoring involves mainly the collection of performance
data for the turbines operating under varying loads and other parameters and using these data to
develop time-dependent and non-linear / linear multi variable models for performance
prediction.
Typical steam turbo generator data
It may be noted that turbine efficiency varies with the connected load as in the example
given above and refers to the tests carried out on a 75-kw condensing steam turbine.
Computer-aided performance monitoring of gas / steam turbine involves the following steps .
1.Turbine data collection. ( energy input / output information )
2.Evaluation of enthalpies of various streams
3.Developing material balance
4.Developing energy balance.
5.Efficiency calculation for the observed conditions.
Load
item 25% 50% 75% 100%
a.Design load kw 75.00
b.Connected load kw 18.75 37.50 55.25 75.00
c.Steam kg/kw 21.61 15.52 12.21 11.39
d.Efficiency % 21.10 30.70 38.20 40.00
e.Steam press kg/cm2 ------------------13.60 ---------------
f.Steam temp oC ------------------ 225 ----------------
g.exhaust press hg mm ---------------- 68.00----------------
h.exhaust temp oC ----------------- 42.00----------------
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
174
6.Deviation Analysis with base data.
7.Identifying causes for corrective action.
Performance of turbines vary with time and connected load as shown in the above
table. As the shaft load increases , efficiency of the turbine increases and the specific
consumption of energy decreases .
It is possible to develop a suitable performance model based on the actual observations.
Necessary flow rate corrections are to be applied for the steam pressure, temperature and flow
rates. Typical output of steam turbine models using the above data is given in box no 4.1. Box
no 4.2 gives the simulated turbine efficiency for new conditions, based on the performance
model developed by this method.
Box 4.1 Steam Turbine Performance Model (load % vs efficiency)
S.E of Model : 0.0001
Box 4.2 Simulated Performance for new conditions.
load% Z1 simulated efficiency
20 20.0000 19.2446
35 35.0000 24.9617
65 65.0000 35.6824
85 85.0000 39.8055
105 105.0000 39.3594
load% X1 efficiency %
observed simulated
25 25.0000 21.1000 21.1000
50 50.0000 30.7000 30.7001
75 75.0000 38.2000 38.1999
100 100.0000 40.0000 40.0000
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
175
Steam Turbine Performance Model 2 ( load vs sp.cons)
load X1 observed simulated sp.con sp.con
25 25.0000 21.6100 21.6100
50 50.0000 15.5200 15.5199
75 75.0000 12.2100 12.2101
100 100.0000 11.3900 11.3900
S I M U L A T E D O U T P U T F O R N E W D A T A
When once the models are developed it is possible to determine the performance level of
each machine with reasonable accuracy at any point of time for various connected load
conditions.
In the case of gas turbines , the above models may be modified to give specific energy
input / kw of net shaft power and efficiency vs load. This technique is superior to the
conventional static performance monitoring method, which does not account for random /
variable parameters.
load Z1 simulated load Z1 simulated sp.cons sp.cons
20 20.0000 23.1873 65 65.0000 13.2190
35 35.0000 18.8217 85 85.0000 11.5996
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
176
Trouble shooting Turbine problems
Turbine problems are similar to compressor problems and may be divided into
mechanical and operational problems.
Mechanical problems include
� Vibration
� Rotor shaft displacement
� Over heating of bearings
� Excessive noise during running
� Impeller failure
� Shaft failure
Operational problems include
� High steam consumption
� Inadequate power generation at full load
� High pressure drop in last stage
� Improper functioning of surface condenser
� Frequent tripping of turbine
� Auxiliary pump failures
Vibration:
Vibration in turbines may be caused not necessarily from turbine section, but it may be due to a
coupled generator.
The turbine has to be decoupled and vibration checked.
In many cases, vibration may be due to poor bearing conditions and / or damaged bearings.
This may also happen due to crevices formed in the rotor shaft at bearing contact sections.
Another reason could be due to rotor imbalance, caused by damage to turbine blades or silica
deposits or products of corrosion.
A systematic analysis must be carried out to determine the cause for vibration.
Vibration trips:
All rotating equipments like turbines, compressors etc are provided with a vibration monitor cum
trip device. When the vibration exceeds a certain allowable limit, the machine will trip
automatically. By analyzing the vibration readings at various sections, it is possible to identify
the section causing vibration and rectify the problem.
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
177
A regular vibration monitoring program may even predict the likely failure of a rotating machine
beforehand.
Rotor shaft displacement:
Reasons for this problem may be due to worn out thrust bearings or crevices in the shaft section
of the bearing housing. When this problem occurs, the bearing temperature will be higher by 40
to 50 oC more than normal. This may be checked in the usual manner.
Over heating of bearings:
Over heating of bearings may be caused by poor / improper lubrication and / or the presence of
wear particles removed from the shaft or bearing. Continuous monitoring of bearing
temperatures, lubrication and lube oil analysis is imperative to check the problem.
Excessive noise during running :
Normally this problem shall be due to rough bearings / shafts or presence of hard materials in the
rotating section. Even if a section of impeller is damaged, this noise may be noticed. Turbine
shut down is necessary to check the problem.
Failure of impellers / shafts :
Thermal shocks, frequent trips, poor material of construction, aging and creep stresses may be
attributed to these failures.
OPERATIONAL PROBLEMS:
High steam consumption :
This could be due to some hindrances in the steam passage. Scales, silica and bacterial / fungus
growth may cause this problem. Besides, clogged impellers, excess friction etc may also cause
this problem.
Inadequate power generation at full load :
For a given steam flow rate at the design temperature and pressure, the power generated must
tally with the turbine characteristic. If the deviation is excessive check the steam quality, flow
rate and the power meter. Also check the RPM of turbine. When all these parameters are normal,
the reason could be with the generator section. Check all the generator parameters like coil
insulation resistance, temperature etc.
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
178
High pressure drop in last stage :
Normally, this happens due to the deposition of silica in the last impeller stage, where the
temperature is low. Silica is converted into silicic acid, which forms iron silicate. This impedes
the flow of steam, besides imbalances in the rotor.
Improper functioning of surface condenser :
This occurs due to poor vacuum in the surface condenser and / of fouling up of the water tubes,
flowing through the condenser for condensing the steam. Check for fouling and poor vacuum in
the SC section by vacuum holding test and tube side pressure drop measurement.
It may be noted that for the same shaft load % , specific consumption of steam or energy
increases with the passage of operating time. This is very reasonable, as turbine blades are
subjected to wear and tear like attrition / erosion , silica deposition , corrosion etc and cause an
imbalance of the rotor. Besides this, the surface condenser also gets fouled up with passage of
time and affects the turbine performance.
box 4.3 Turbine Performance with time
.No of data sets used in the model : 5
Independant variables used in the model : 2
Variables used in the model are ......
variable 1 is load%
variable 2 is year
variable 3 is sp.con kg/kw
load% year sp.con
kg/kw
25.0 1.0 21.61
50.0 1.0 15.52
75.0 1.0 12.21
100.0 1.0 11.39
25.0 2.0 22.55
50.0 2.0 16.52
75.0 2.0 13.35
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
179
100.0 2.0 12.51
25.0 4.0 23.00
50.0 4.0 17.20
75.0 4.0 14.10
100.0 4.0 13.30
Model is of the form
sp.con kg/kw = 100.3775 * (load %) -0.47874 * (op.year) 0.06896
Standard Error of the model = 0.30353
SIMULATED OUTPUT FOR NEW INPUT PARAMETERS
LOAD % OPTG SP.CONS YR KG/KW
90.0 4.5 12.91545
70.0 4.5 14.56670
65.0 4.5 15.09278
55.0 4.5 16.34941
40.0 4.5 19.04203
35.0 4.5 20.29907
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
180
Loss Control Related to Power Plants
Introduction:In power plant operation physical losses occur in fuel section and steam circuit during
various stages of operation. The first step is to identify the loss areas and quantify the losses for
control action. In power plant operation, factors attributing to the losses can be classified as :
Coal handling losses during transportation
Losses due to coal dust generation
Spillage
Day to day quality variation
Measurement
Steam transportation losses
Safety valve popping
ID / FD fan losses etc.
Plant Losses Break-up:
Table 3.01 shows a typical break-up of various losses starting from coal transportation.
This will vary from plant to plant and the type of coal handling and transportation methods, coal
conveyor maintenance efficiency of the o&m personnel etc.
This type of loss break up may be carried out for any coal based power plant for loss
evaluation and control.
This exercise will help the power plant personnel refiners in taking appropriate
corrective action for loss control. As could be seen from the loss analysis table shown above,
maximum loss iin the refinery occurs in the form of process loss amounting to 69 % of the total
loss. For this particular system under consideration, loss control priorities must be set in the
following order.
Coal Handling Area
Coal receipt / dispatch / transfer
Coal crushing section
Pulverization section
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
181
Coal fluidization
Combustion
Carry over in fly ash
Carbon in refuse
Carbon Losses in boiler operation :
Wagons
As may be seen from the typical loss figures in various section, the quantity of coal
reaching the boiler as pulverized coal = 0.99 x 0. 0.995 x 0.995 x 0.995= 0.9752 / unit weight of
raw coal.
Entire pulverized fuel is not burnt in the boiler to complete combustion. When
investigating steam systems, the boiler is one of the primary targets for energy-efficiency
improvement. There are many tools used in the evaluation and management of boiler
performance. One of the most useful tools is boiler efficiency. Boiler efficiency describes the
fraction of fuel energy that is converted into useful steam energy. Of course, the fuel input
energy that is not converted into useful steam energy represents the losses of the boiler operation.
Coal receipt areaSpillage during receipt 0.5 to 2.0 % on billed Quantity
Transfer to crusher house by conveyors
Transfer loss 0.2 to 0.5%
Crushingoperation
Losses during crushing0.5 to 0.6 %
PulverizationLosses 0.5 to 1.0 %
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
182
Boiler investigations generally examine the losses by identifying the avenues of loss, measuring
the individual loss and developing a strategy for loss reduction.
There are many avenues of loss encountered in boiler operations. Typically, the dominant
loss is associated with energy leaving the boiler with combustion gases. The temperature of
exhaust gases is an indication of their energy content.
Ensuring that the heat-transfer surfaces of the boiler are clean is a major point of focus
for managing the thermal energy in exhaust gases. Energy can be recovered from exhaust gases
by transferring thermal energy from the high-temperature gases to boiler feed water, or to the
combustion air entering the boiler.
Another aspect of exhaust-gas energy management, which is the focus here, is
combustion management. It should be noted that the temperature- and combustion-related
attributes of exhaust gases are interrelated – they combine to represent the stack loss of the
boiler. Again, this is typically the dominant loss for the boiler. Stack loss is dependent on the
operating characteristics of the boiler, the equipment installed and the type of fuel burned in the
boiler. Stack loss generally ranges from as much as 30 percent for a green-wood-fired boiler, to
18 percent for a typical natural-gas-fired boiler, to 12 percent for an oil-fired boiler, to as low as
9 percent for a coal-fired boiler.
It must be pointed out that the stack-loss range is wide for any given fuel. To address
your question, we will examine the combustion of a simple fuel – methane (CH4). The chemical
equation for the reaction of methane with oxygen (O2) is presented below:
CH4 + 2O2 CO2 + 2H2O
The combustion process does not proceed in a perfect manner. A fuel molecule may
encounter less oxygen than is required for complete combustion. The result will be partial
combustion; the exhaust gases will then contain some unreacted fuel and some partially reacted
fuel. Generally, these unburned fuel components are in the form of carbon monoxide (CO),
hydrogen (H2) and other fuel components that may include the fully unreacted fuel source,
which in this case is methane.
When unburned fuel is found to be part of the combustion products, a portion of the fuel
that was purchased is consequently discharged from the system, unused. It is also important to
note that unburned fuel can accumulate to a point where a safety hazard could result. Unburned
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
183
fuel can burn in a part of the boiler not designed for combustion – under certain conditions, the
materials can even explode.
Pump and Compressor seal leakage
Normally gland packing or mechanical seals are used but leakage through packing is inevitable
while the leakage through Mechanical seal is almost nil. Seal leakage will vary significantly
with operating pressure, gas properties and condition of the seals. To estimate the gland leak,
sample from the leakage source is collected in a graduated cylinder for a known period of time
and oil content estimated by laboratory analysis or visually. Then the leakage rate is calculated
over a period on one hr. When this exercise is carried out periodically, it is possible to determine
the process loss from leakage. A typical Format is given below for determination of leakage
loss.
Format 1. Leakage Determination from Plant Area
No Unit Pump /
Comp
time
mt /sec
Volum
e in cc
Density
in
gm/ml
Loss
kg/hr
Loss is calculated by the following formula. Loss in kg/hr = ( Vol in cc / time in mt ) * 60 * density * 1/1000Based on the loss data, maintenance priorities could be accorded. From a historical data of leaks, it is possible to identify the source of frequent leaks and causes for remedial action.Example:
1 BFW BFW! 30 325 1.0000 22.4246
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
184
Loss from BFW1 (325/30)*3600*1.0000*1/1000 = 39.0 kg/h Yearly loss @ 8000 hrs of operation = 312.0 tpy
Loss Monitoring by Models:
From the historic data maintained above, it is possible to develop a hydrocarbon loss
scenario and it's impact on profitability if the situation remains unattended. This approach
could be used for economic justification of equipment replacement / maintenance. A
typical data input and hydrocarbon loss prediction is given in table 3.03 and fig 3.02 &
3.03.
Table 3.03 Leakage Determination by NL Model
month X1 ACTUAL ESTIMATED leak kg/h leak kg/h
1 1.0000 0.3500 0.6975 2 2.0000 1.8900 2.8464 3 3.0000 5.4700 4.9154 4 4.0000 8.8600 6.9043 6 6.0000 10.2100 10.6422 8 8.0000 13.1200 14.0601 10 10.0000 16.7100 17.1580 12 12.0000 20.5500 19.9360
STANDARD ERROR OF ESTIMATE : 0.9231
Leakage forecast using NL Model
month Z1 ESTIMATED leak kg/h
14 14.0000 22.3939 16 16.0000 24.5318
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
185
18 18.0000 26.3498
Loss during strainer cleaning
Strainer cleaning is a periodical job and during this operation a sizeable amount is being
drained either to ETP or open drain. Although theoretically it should go to ETP/CBD, in
practice it goes to surface drain. Hence, care should be taken during cleaning of the
strainer so that drained material should go to ETP/CBD and ensure no material escapes to
the surface drain. This leads to environmental pollution, safety hazard and also loss of
hydrocarbons by evaporation. This could be ensured by the analysis of surface drain
water for oil content. A format could be developed in the same lines as above to
determine losses due to draining. A historical data on oil content will reveal the operation
and maintenance efficiency of the refinery.
Typical values for well maintained seals
100 SCF/ W per rod for reciprocating comp.
8.5 1b/day per seal for centrifugal comp.
4.0 1b/day per seal for pump
These are representative average numbers only and actual rates at a given location for a
given pump or compressor will vary considerably. Replacement of packing by good
quality mechanical seal will reduce the losses. It will also reduce the power consumption
of the drive and hence energy savings. Failure of seals contribute to losses which can be
minimized by prompt rectification / replacement of seals.
Drip Losses:
Drips and leaks add to high losses. A hydrocarbon leak is easily detectable. Necessary
steps should therefore, be taken to stop the same without any loss of time. The table 3.03
shows the amount of leakage loss in litres (approx.)
Table 3.03 . Estimation of Leakage Loss due to drips & leaks
Leakage In one
minute
In one
hour (Lts)
In one
day (Lts)
In one
week
In one
month (Lts)
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
186
(Lts.) (Lts)
1. Drop/sec. 0.003 0.17 4.4 31.0 132
2. Drop/sec. 0.009 0.57 14.6 101.0 398
A drop breaking
into
a stream
0.057 3.89 93.4 634 2705
1/8" Stream 0.65 43.0 1011 6842 29594
3/16" Stream 1.11 70.0 1591 11456 49165
1/4" Stream 2.36 152 3500 248211 106445
Leakage of product not only cost excise duty & product loss but also is a fire hazard.
Leakage in a refinery usually occur at :
Pipe joints
Flanges
Valve glands
Pump glands
Storage tanks
Loading rack
PREVENTION AND REDUCTION OF LEAKAGE LOSSES
a) Leakage through pipe joints
As and when a leakage is detected through a threaded pipe joint, the pipe line should be
immediately repaired. Expansion joint should be provided on long pipeline. Thermal
expansion relief provision must be made. This will take the undue stress and strain of the
pipelines due to variation of day's temperature.
b) Leakage through Flanges
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
187
Flanges should be of standard thickness. The nuts should be tightened crosswise and
diagonally opposite. All nuts should be tightened little by little cross-wise to ensure
equal tightening. Proper gasket packing with liberal use of grease is to be used. Packing
should not extend beyond the flange collar. This helps in tightening the holes evenly all
round.
c) Leakage through valve glands
For globe valves the gland packing should be asbestos rope packing.
For gate valves and valves of bigger diameter where the gap between the valve
stem and the body is more than 3 mm, graphite asbestos rope packing of square
cross section should be used.
The old packing must be removed from the valve and then replace with the new
packing adequately. The gland nuts should be tightened evenly.
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
188
Controlling Steam losses by survey Steam losses occur due to leaks from gaskets, pipe flanges, pin holes, steam trap passing
etc. This may be quantified and exact reason identified for correcting the situation and
minimizing steam leaks. A typical output is given below.
=======================================================
S T E A M L E A K E S T I M A T I O N
ABC Oil Corporation Ltd
=======================================================
User : xxx Refinery
Unit : India
Input file : C:\tthermw\stmleak.in
Output file : C:\tthermw\stmleak.out
Archive ,, : c:\tthermw\stmleak.arc
Date : 01-22-2008\05:54:44
1. P R O G R A M
O U T P U T UNIT TAG LEAKAGE PLUME SIZE STEAM Loss CODE NO CATEGORY (metre) crude 1 a 1.55 44.826 crude 2 b 1.72 61.341 crude 3 a 1.55 44.826 crude 4 b 1.72 61.341 crude 5 a 1.5 40.876 crude 6 b 1.02 16.860 crude 7 a 1.34 30.428 crude2 2 b 0.72 9.693 crude2 1 a 1.25 25.772 crude2 2 b 1.22 24.385 crude2 1 a 1.15 21.430 crude2 2 b 1.02 16.860 crude2 1 a 1.15 21.430 VBU 2 b 1.22 24.385 VBU 1 a 1.15 21.430 VBU 2 b 1.42 35.267
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
189
2.TOTAL STEAM LEAK IDENTIFIED
Total Steam Leaks in kg/h 501.1516
Equivalent Fuel in kg/h 7099.648
Steam Enthalpy for estimation (kcal/kg) : 10200
Fuel Calorific Value in kcal/kg : 720
3. LEAK CATEGORISATION
Code 'a' denotes leaks from pin holes & weld joints
Code 'b' denotes leaks from valve glands.
Code 'c' denotes gasket leak from joints.
Code 'd' denotes gasket miscellaneous leaks
***** end of steam leak program *****
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
160
4
PERFORMANCE MONITORING TECHNIQUES
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
161
4. Performance Monitoring Techniques
1. Monitoring Boiler Efficiency / Performance :
Boiler efficiency is determined by operation and maintenance practices of the total
system. Efficiency is also an indicator of mechanical condition of the boiler system as a whole.
Hence if the boiler efficiency is regularly monitored, it is possible to identify deterioration trend
and take corrective action. Using the deterioration trend, it is possible to predict the boiler
efficiency at a future period, by using historical models.
For boiler efficiency monitoring purposes either direct of indirect method or the
average may be selected and consistently used. A typical output of the boiler efficiency
program was given in the earlier chapter.
Precautions :Certain precautions are to be followed when boiler efficiency is determined as
listed below.
o Boiler load must be consistent.
o All flow meters must be calibrated and errors minimized.
o Pressure , temperature and draft profiles in the boiler circuit must be kept consistent
during the run.
While fig 4.01 gives a typical boiler efficiency curve as a function of load , impact of
running hrs on the efficiency is also given in the figure for various load conditions . Box 4.02
gives the performance models at base case and after ten months of operating time which may
be used for evaluation of performance and timely maintenance action.
Differentiating the above functions w.r.t load % and equating to 0 it may be found
that the maximum achievabe efficiency in the base case is 89.80 at 103.3% of design load ,
whereas the maximum achievable efficiency at the end of ten months of operation is 84.00 % at
a load of 99.16% .This clearly indicates , there is a drop in the efficiency of boiler with
the passage of time and load% .
This tends to affect the operating cost of the boiler and it is possible to determine
the optimum cycle length at which the boiler has to be shut down and cleaned so that the
efficiency may be restored to the level of base case .
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
162
Fig 4.01. Boiler Efficiency vs Load %
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
163
Performance monitoring technique:
Efficiency of boilers must be calculated for each boiler periodically and recorded
stream day wise. All the input parameters must be as accurate as possible and may be used as an
effective management tool to determine
o What is the current efficiency of the boiler ?
o What should be the efficiency at the current stream day and load %
o Whether the efficiency is within tolerable limits ?
o If not what are the reasons for lower efficiency ?
o What are the actual losses against base case ?
o Should the boiler be operated in the existing condition or should it be repaired ?
o Whether the boiler can be maintained or replaced etc.
A typical decision flow diagram to arrive at the right decision is given in fig 4.05 .
This methodology could be arrived at for any number of boilers when once the models are
developed for each controllable parameter .This uses a number of models for arriving at the
right decision.
Fig 4.05. Boiler Efficiency
management
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
164
Typical boiler efficiency program output
=================================================================
RESULTS - BOILER EFFICIENCY
=================================================================
User : ABC Refinery
Unit : India
Equipment : blr1
Input file : c:\tthermw\blr1.in
Output files: c:\tthermw\blr1.out
Output file2: c:\tthermw\blr1.mv
Archive file: c:\tthermw\boiler.arc
Date of Obs : 12.10.07
Stream Day : 100
Run Date : Run No : 1
I.Fuel Data
Carbon .87000
Hydrogen .12000
Moisture .00000
Oxygen .00000
Sulfur .01000
Nitrogen .00000
Ash .00000
II.Observed Process parameters
Fired Duty mmkcal/h 65.0000
% Heat load 100.0000
% Unburnt matter in refuse 0.0000
Amb.Temperature oC 30.0000
Flue Gas Temp oC 180.0000
Relative Humidity 0.0000
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
165
Excess Air 15.6454
Fuel High Heating Value - Kcal/Kgm 11163.8896
Fuel Low Heating Value - Kcal/kgm 10515.8887
III.Energy Losses
a.Dry Gas Loss : 5.2683
b.Air Moisture loss : 0.0000
c.Combustion Moisture Loss : 6.4719
d.Fuel Moisture Loss : 0.0000
e.Radiation Loss : 1.0000
f.Blow Down Loss : 2.0000
g.Unaccounted Loss : 0.3100
g.Loss due to combustibles : 0.0000
Total Loss % ( dry basis) 15.0502
IV.Boiler Efficiency
a.EFFICIENCY HHV basis : 84.9498
b.EFFICIENCY LHV basis : 90.1845
NOTE:
a.Air misture loss is due to moisture present in combustion air.
b.Combustion moisture is due to combustion of H2 in fuel to water.
c.Fuel moisture is due to presence of water in the fuel fired.
d.Radiation loss is due to heat loss from the exposed boiler surface.
e.Blow down is energy equivalent as % of fuel calorifiv value.
*** end of boiler efficiency program (indirect method) ***
Output information may be used to analyze performance of each boiler for efficiency
improvement.
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
166
A typical energy loss break-up for the boiler is given in fig 4.06. From this it is
obvious that all losses except wet losses show an increasing trend with respect to time.
Controllable losses in this case are dry gas loss , convection and radiation losses and blow
down loss .
Fig 4.06. Energy loss break up from boiler efficiency program output.
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
167
Observations:
o Increase in dry gas loss indicates gradual increase in excess air which could be due to
air leaks in the shell section . This is normally diagnosed by oxygen analysis at
combustion , pre-convection and post convection zones.
o Increasing trend in convection and radiation losses is indicated by high skin
temperature of the shell. This situation occurs due to thinning of the insulation brick. It
is a good practice to monitor this loss known as setting loss periodically, for deciding
the optimal replacement time.
o High blow-down rate could be due to poor operation practice or poor boiler feed
water quality . This may be checked from the TDS levels of make-up water , drum
water and allowable TDS limits.
Systematic recording and review of boiler parameters is very imperative to determine the
energy efficiency of a boiler.
A typical analysis data for a set of boilers is given in box 4.03. This is actually the
program output of boiler efficiency monitoring system program using a popular program,
which utilizes indirect method of determining boiler efficiency.
Blow down rate as % of energy input must be added to the last column to determine the
efficiency of boilers listed below.
Boiler Efficiency Models
Since boiler efficiency is determined by a number of parameters , it is imperative to
identify the impact of all parameters by suitable scientific methods. Boiler efficiency is
invariably affected by running days and load factor.
Scaling increases with the passage of time on both the water side and tube side. This
is reflected by rise in water tube skin temperature, flue gas temperature etc .
Since the boiler load is always dynamic depending on steam demand, the efficiency
could be estimated as a function of operating period and load % involving two variables .
Even a linear multi variable model is good enough for monitoring purpose. For this
purpose, common programs like MS Excel may be used .
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
168
Performance Monitoring of boilers
Performance monitoring of boilers is very imperative in power and process industries
as the major chunk of cost is governed by the fuel consumed in the boilers .
Since fuels in boilers are not in their pure form , they tend to deposit soot , scales etc in
the economizer , super heater , water tubes etc which retard heat transfer efficiency with the
passage of time .
In addition to this , steady loss in evaporating capacity is also experienced due to
accumulation of scales on the water side with the passage of time . The decrease in heat-
transfer rate is a function of time and is given by
where
U is the heat transfer coefficient in kcal /hr / m2 /oC
t is the time lapsed since last cleaning.
For minimizing the scale deposition on the convective section of the boiler, soot blowers
are used periodically to remove the loose deposits .However, hard scales tend to deposit with the
passage of time .
Hence performance monitoring of the boiler is essential to decide at what point of time
the boiler has to be taken out of service for cleaning and other maintenance jobs.
Like the case of heaters, boiler efficiency also varies with the passage of time due to the
above factors. In addition to this , boiler capacity utilization also affects the efficiency and
boiler manufacturers provide the boiler operating characteristics data.
Hence the performance of the boiler may be evaluated by the observed efficiency and
compared to the basic characteristics for the same operating conditions. Any abnormal
deviations observed could therefore be investigated to pin-point the problem area for
necessary action.
It is obvious from the above the boiler efficiency can be managed by monitoring the
efficiency of economizer, air pre heater and super heater by appropriate techniques.
1 / U2 = A + B * t
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
169
Specific Fuel Consumption
Specific fuel consumption is normally assumed to be a performance Indicator of energy
efficiency of the system. The system could be a heater or boiler or turbine. In conventional
Energy Management, Specific Energy Consumption is defined as the quantity of fuel consumed
in kgs or Lbs as given below.
Equipment Specific Fuel Consumption
1.Heater Kg fuel / Ton of feed processed
2.Boiler ton fuel / Ton steam produced.
3.Gas Turbine ton fuel / MW power generated.
4.Steam Turbine ton fuel eqvt steam / MW of Power
generated.
Examples given below are indicative of the concept of specific fuel consumption.
Specific fuel consumption may be used as a thumb rule to determine the equipment performance
/ efficiency by a quick glance. When all the parameters remain constant, specific fuel
consumption is definitely a simple and good performance indicator of the system under
consideration.
Item Feed tons Fuel tons Sp.Fuel
Consumption Unit
1.Heater 500 7.5 15 Kg/ton of
feed
2.Boiler 150 10 0.06667 Ton/ton of
steam prodn
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
170
3.Gas Turbine 25 MW 4.50 0.180 Tons of fuel /
MW
4.Steam
Turbine
10 MW 100 tons
steam
10 Tons steam /
mw
Historical data of specific energy consumption of system / systems in an indicator of
efficiency. With the passage of time efficiency starts dropping down, thereby increasing the
specific consumption of fuels.
Factors Affecting Specific Fuel Consumprion :
Fuel is consumed only in process heaters, boilers and Gas Turbines in the refining
industry. A number of process parameters affect specific fuel consumption in the case of heater /
boiler. They are
Capacity utilization and
Heater operating parameters such as
Draft
Transfer Temperature
Fuel type and it's Calorific Value
Fuel Mix ( % Oil , % Gas ) and
Combustion Efficiency of Fuel
Impact of capacity utilization on a fired heater efficiency was shown in earlier chapter when
all the other parameters are constant. Combustion Efficiency is a function of fuel mix used in a
heater / boiler.
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
171
Performance Monitoring of turbines.
Performance and condition monitoring of turbines are two different areas of energy
efficiency management. While condition monitoring refers to the mechanical condition of the
rotating equipment and is concerned with vibration level, bearing conditions , mechanical
condition etc. performance monitoring refers to the energy efficiency of the turbine under
consideration.
But it is necessary to know that condition of the rotating equipment has direct impact on
energy consumption and efficiency of the machine.The physical layout of various turbines were
presented in earlier chapters.
Performance Monitoring of steam turbines
While turbine efficiency is calculated based on the input/output principles, performance
of turbines vary with the load factor and aging / on -stream hours / days. Turbine
manufacturers often supply the performance characteristics of each machine in the form of
a curve for various energy input conditions.
Actual performance of the turbine may be compared with the base case to determine
whether or not the efficiency is within the acceptable limits. A typical steam turbine
characteristic is given in fig 4.5(a).
This characteristic curve gives the input steam rate vs output in kw for any chosen
extraction steam rate.
The graph is developed from the actual performance data for a small machine. For
performance monitoring of the turbine , the output observed at any load is compared against
the base case ( in this case characteristics curve) and checked whether the performance is
normal or not.
Case given in fig 4.5(a) has extraction rates of 0 ,5 ,10 and 15 t/hr and any intermediate
value is interpolated.
An example of how the turbine characteristics may be used for performance monitoring
is given in the section.
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
172
fig 4.5(a) .Typical Steam turbine characteristics.
observed data: steam inlet 20.8 t/h
pass - out 10.0 t/h
power output 2100 kw
From the characteristics, power output should be 2750 kw.
Observed Deviation in performance is -650 kw
% deviation from base value = (-650/2750)* 100
= 23.60%
The deviation is very high and needs a thorough investigation. If the turbine operation
continues, excess steam consumption would have been about 3.0 t/hr for the same output. (
given by y-x )
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
173
Increase in annual steam consumption based on 8000 hrs of operation works out to
24000 tons. For applying this monitoring technique, user must have the complete turbine
characteristic for each machine and refer to the same every time the checking is required.
Computer-aided performance monitoring involves mainly the collection of performance
data for the turbines operating under varying loads and other parameters and using these data to
develop time-dependent and non-linear / linear multi variable models for performance
prediction.
Typical steam turbo generator data
It may be noted that turbine efficiency varies with the connected load as in the example
given above and refers to the tests carried out on a 75-kw condensing steam turbine.
Computer-aided performance monitoring of gas / steam turbine involves the following steps .
1.Turbine data collection. ( energy input / output information )
2.Evaluation of enthalpies of various streams
3.Developing material balance
4.Developing energy balance.
5.Efficiency calculation for the observed conditions.
Load
item 25% 50% 75% 100%
a.Design load kw 75.00
b.Connected load kw 18.75 37.50 55.25 75.00
c.Steam kg/kw 21.61 15.52 12.21 11.39
d.Efficiency % 21.10 30.70 38.20 40.00
e.Steam press kg/cm2 ------------------13.60 ---------------
f.Steam temp oC ------------------ 225 ----------------
g.exhaust press hg mm ---------------- 68.00----------------
h.exhaust temp oC ----------------- 42.00----------------
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
174
6.Deviation Analysis with base data.
7.Identifying causes for corrective action.
Performance of turbines vary with time and connected load as shown in the above
table. As the shaft load increases , efficiency of the turbine increases and the specific
consumption of energy decreases .
It is possible to develop a suitable performance model based on the actual observations.
Necessary flow rate corrections are to be applied for the steam pressure, temperature and flow
rates. Typical output of steam turbine models using the above data is given in box no 4.1. Box
no 4.2 gives the simulated turbine efficiency for new conditions, based on the performance
model developed by this method.
Box 4.1 Steam Turbine Performance Model (load % vs efficiency)
S.E of Model : 0.0001
Box 4.2 Simulated Performance for new conditions.
load% Z1 simulated efficiency
20 20.0000 19.2446
35 35.0000 24.9617
65 65.0000 35.6824
85 85.0000 39.8055
105 105.0000 39.3594
load% X1 efficiency %
observed simulated
25 25.0000 21.1000 21.1000
50 50.0000 30.7000 30.7001
75 75.0000 38.2000 38.1999
100 100.0000 40.0000 40.0000
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
175
Steam Turbine Performance Model 2 ( load vs sp.cons)
load X1 observed simulated sp.con sp.con
25 25.0000 21.6100 21.6100
50 50.0000 15.5200 15.5199
75 75.0000 12.2100 12.2101
100 100.0000 11.3900 11.3900
S I M U L A T E D O U T P U T F O R N E W D A T A
When once the models are developed it is possible to determine the performance level of
each machine with reasonable accuracy at any point of time for various connected load
conditions.
In the case of gas turbines , the above models may be modified to give specific energy
input / kw of net shaft power and efficiency vs load. This technique is superior to the
conventional static performance monitoring method, which does not account for random /
variable parameters.
load Z1 simulated load Z1 simulated sp.cons sp.cons
20 20.0000 23.1873 65 65.0000 13.2190
35 35.0000 18.8217 85 85.0000 11.5996
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
176
Trouble shooting Turbine problems
Turbine problems are similar to compressor problems and may be divided into
mechanical and operational problems.
Mechanical problems include
� Vibration
� Rotor shaft displacement
� Over heating of bearings
� Excessive noise during running
� Impeller failure
� Shaft failure
Operational problems include
� High steam consumption
� Inadequate power generation at full load
� High pressure drop in last stage
� Improper functioning of surface condenser
� Frequent tripping of turbine
� Auxiliary pump failures
Vibration:
Vibration in turbines may be caused not necessarily from turbine section, but it may be due to a
coupled generator.
The turbine has to be decoupled and vibration checked.
In many cases, vibration may be due to poor bearing conditions and / or damaged bearings.
This may also happen due to crevices formed in the rotor shaft at bearing contact sections.
Another reason could be due to rotor imbalance, caused by damage to turbine blades or silica
deposits or products of corrosion.
A systematic analysis must be carried out to determine the cause for vibration.
Vibration trips:
All rotating equipments like turbines, compressors etc are provided with a vibration monitor cum
trip device. When the vibration exceeds a certain allowable limit, the machine will trip
automatically. By analyzing the vibration readings at various sections, it is possible to identify
the section causing vibration and rectify the problem.
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
177
A regular vibration monitoring program may even predict the likely failure of a rotating machine
beforehand.
Rotor shaft displacement:
Reasons for this problem may be due to worn out thrust bearings or crevices in the shaft section
of the bearing housing. When this problem occurs, the bearing temperature will be higher by 40
to 50 oC more than normal. This may be checked in the usual manner.
Over heating of bearings:
Over heating of bearings may be caused by poor / improper lubrication and / or the presence of
wear particles removed from the shaft or bearing. Continuous monitoring of bearing
temperatures, lubrication and lube oil analysis is imperative to check the problem.
Excessive noise during running :
Normally this problem shall be due to rough bearings / shafts or presence of hard materials in the
rotating section. Even if a section of impeller is damaged, this noise may be noticed. Turbine
shut down is necessary to check the problem.
Failure of impellers / shafts :
Thermal shocks, frequent trips, poor material of construction, aging and creep stresses may be
attributed to these failures.
OPERATIONAL PROBLEMS:
High steam consumption :
This could be due to some hindrances in the steam passage. Scales, silica and bacterial / fungus
growth may cause this problem. Besides, clogged impellers, excess friction etc may also cause
this problem.
Inadequate power generation at full load :
For a given steam flow rate at the design temperature and pressure, the power generated must
tally with the turbine characteristic. If the deviation is excessive check the steam quality, flow
rate and the power meter. Also check the RPM of turbine. When all these parameters are normal,
the reason could be with the generator section. Check all the generator parameters like coil
insulation resistance, temperature etc.
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
178
High pressure drop in last stage :
Normally, this happens due to the deposition of silica in the last impeller stage, where the
temperature is low. Silica is converted into silicic acid, which forms iron silicate. This impedes
the flow of steam, besides imbalances in the rotor.
Improper functioning of surface condenser :
This occurs due to poor vacuum in the surface condenser and / of fouling up of the water tubes,
flowing through the condenser for condensing the steam. Check for fouling and poor vacuum in
the SC section by vacuum holding test and tube side pressure drop measurement.
It may be noted that for the same shaft load % , specific consumption of steam or energy
increases with the passage of operating time. This is very reasonable, as turbine blades are
subjected to wear and tear like attrition / erosion , silica deposition , corrosion etc and cause an
imbalance of the rotor. Besides this, the surface condenser also gets fouled up with passage of
time and affects the turbine performance.
box 4.3 Turbine Performance with time
.No of data sets used in the model : 5
Independant variables used in the model : 2
Variables used in the model are ......
variable 1 is load%
variable 2 is year
variable 3 is sp.con kg/kw
load% year sp.con
kg/kw
25.0 1.0 21.61
50.0 1.0 15.52
75.0 1.0 12.21
100.0 1.0 11.39
25.0 2.0 22.55
50.0 2.0 16.52
75.0 2.0 13.35
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
179
100.0 2.0 12.51
25.0 4.0 23.00
50.0 4.0 17.20
75.0 4.0 14.10
100.0 4.0 13.30
Model is of the form
sp.con kg/kw = 100.3775 * (load %) -0.47874 * (op.year) 0.06896
Standard Error of the model = 0.30353
SIMULATED OUTPUT FOR NEW INPUT PARAMETERS
LOAD % OPTG SP.CONS YR KG/KW
90.0 4.5 12.91545
70.0 4.5 14.56670
65.0 4.5 15.09278
55.0 4.5 16.34941
40.0 4.5 19.04203
35.0 4.5 20.29907
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
180
Loss Control Related to Power Plants
Introduction:In power plant operation physical losses occur in fuel section and steam circuit during
various stages of operation. The first step is to identify the loss areas and quantify the losses for
control action. In power plant operation, factors attributing to the losses can be classified as :
Coal handling losses during transportation
Losses due to coal dust generation
Spillage
Day to day quality variation
Measurement
Steam transportation losses
Safety valve popping
ID / FD fan losses etc.
Plant Losses Break-up:
Table 3.01 shows a typical break-up of various losses starting from coal transportation.
This will vary from plant to plant and the type of coal handling and transportation methods, coal
conveyor maintenance efficiency of the o&m personnel etc.
This type of loss break up may be carried out for any coal based power plant for loss
evaluation and control.
This exercise will help the power plant personnel refiners in taking appropriate
corrective action for loss control. As could be seen from the loss analysis table shown above,
maximum loss iin the refinery occurs in the form of process loss amounting to 69 % of the total
loss. For this particular system under consideration, loss control priorities must be set in the
following order.
Coal Handling Area
Coal receipt / dispatch / transfer
Coal crushing section
Pulverization section
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
181
Coal fluidization
Combustion
Carry over in fly ash
Carbon in refuse
Carbon Losses in boiler operation :
Wagons
As may be seen from the typical loss figures in various section, the quantity of coal
reaching the boiler as pulverized coal = 0.99 x 0. 0.995 x 0.995 x 0.995= 0.9752 / unit weight of
raw coal.
Entire pulverized fuel is not burnt in the boiler to complete combustion. When
investigating steam systems, the boiler is one of the primary targets for energy-efficiency
improvement. There are many tools used in the evaluation and management of boiler
performance. One of the most useful tools is boiler efficiency. Boiler efficiency describes the
fraction of fuel energy that is converted into useful steam energy. Of course, the fuel input
energy that is not converted into useful steam energy represents the losses of the boiler operation.
Coal receipt areaSpillage during receipt 0.5 to 2.0 % on billed Quantity
Transfer to crusher house by conveyors
Transfer loss 0.2 to 0.5%
Crushingoperation
Losses during crushing0.5 to 0.6 %
PulverizationLosses 0.5 to 1.0 %
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
182
Boiler investigations generally examine the losses by identifying the avenues of loss, measuring
the individual loss and developing a strategy for loss reduction.
There are many avenues of loss encountered in boiler operations. Typically, the dominant
loss is associated with energy leaving the boiler with combustion gases. The temperature of
exhaust gases is an indication of their energy content.
Ensuring that the heat-transfer surfaces of the boiler are clean is a major point of focus
for managing the thermal energy in exhaust gases. Energy can be recovered from exhaust gases
by transferring thermal energy from the high-temperature gases to boiler feed water, or to the
combustion air entering the boiler.
Another aspect of exhaust-gas energy management, which is the focus here, is
combustion management. It should be noted that the temperature- and combustion-related
attributes of exhaust gases are interrelated – they combine to represent the stack loss of the
boiler. Again, this is typically the dominant loss for the boiler. Stack loss is dependent on the
operating characteristics of the boiler, the equipment installed and the type of fuel burned in the
boiler. Stack loss generally ranges from as much as 30 percent for a green-wood-fired boiler, to
18 percent for a typical natural-gas-fired boiler, to 12 percent for an oil-fired boiler, to as low as
9 percent for a coal-fired boiler.
It must be pointed out that the stack-loss range is wide for any given fuel. To address
your question, we will examine the combustion of a simple fuel – methane (CH4). The chemical
equation for the reaction of methane with oxygen (O2) is presented below:
CH4 + 2O2 CO2 + 2H2O
The combustion process does not proceed in a perfect manner. A fuel molecule may
encounter less oxygen than is required for complete combustion. The result will be partial
combustion; the exhaust gases will then contain some unreacted fuel and some partially reacted
fuel. Generally, these unburned fuel components are in the form of carbon monoxide (CO),
hydrogen (H2) and other fuel components that may include the fully unreacted fuel source,
which in this case is methane.
When unburned fuel is found to be part of the combustion products, a portion of the fuel
that was purchased is consequently discharged from the system, unused. It is also important to
note that unburned fuel can accumulate to a point where a safety hazard could result. Unburned
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
183
fuel can burn in a part of the boiler not designed for combustion – under certain conditions, the
materials can even explode.
Pump and Compressor seal leakage
Normally gland packing or mechanical seals are used but leakage through packing is inevitable
while the leakage through Mechanical seal is almost nil. Seal leakage will vary significantly
with operating pressure, gas properties and condition of the seals. To estimate the gland leak,
sample from the leakage source is collected in a graduated cylinder for a known period of time
and oil content estimated by laboratory analysis or visually. Then the leakage rate is calculated
over a period on one hr. When this exercise is carried out periodically, it is possible to determine
the process loss from leakage. A typical Format is given below for determination of leakage
loss.
Format 1. Leakage Determination from Plant Area
No Unit Pump /
Comp
time
mt /sec
Volum
e in cc
Density
in
gm/ml
Loss
kg/hr
Loss is calculated by the following formula. Loss in kg/hr = ( Vol in cc / time in mt ) * 60 * density * 1/1000Based on the loss data, maintenance priorities could be accorded. From a historical data of leaks, it is possible to identify the source of frequent leaks and causes for remedial action.Example:
1 BFW BFW! 30 325 1.0000 22.4246
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
184
Loss from BFW1 (325/30)*3600*1.0000*1/1000 = 39.0 kg/h Yearly loss @ 8000 hrs of operation = 312.0 tpy
Loss Monitoring by Models:
From the historic data maintained above, it is possible to develop a hydrocarbon loss
scenario and it's impact on profitability if the situation remains unattended. This approach
could be used for economic justification of equipment replacement / maintenance. A
typical data input and hydrocarbon loss prediction is given in table 3.03 and fig 3.02 &
3.03.
Table 3.03 Leakage Determination by NL Model
month X1 ACTUAL ESTIMATED leak kg/h leak kg/h
1 1.0000 0.3500 0.6975 2 2.0000 1.8900 2.8464 3 3.0000 5.4700 4.9154 4 4.0000 8.8600 6.9043 6 6.0000 10.2100 10.6422 8 8.0000 13.1200 14.0601 10 10.0000 16.7100 17.1580 12 12.0000 20.5500 19.9360
STANDARD ERROR OF ESTIMATE : 0.9231
Leakage forecast using NL Model
month Z1 ESTIMATED leak kg/h
14 14.0000 22.3939 16 16.0000 24.5318
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
185
18 18.0000 26.3498
Loss during strainer cleaning
Strainer cleaning is a periodical job and during this operation a sizeable amount is being
drained either to ETP or open drain. Although theoretically it should go to ETP/CBD, in
practice it goes to surface drain. Hence, care should be taken during cleaning of the
strainer so that drained material should go to ETP/CBD and ensure no material escapes to
the surface drain. This leads to environmental pollution, safety hazard and also loss of
hydrocarbons by evaporation. This could be ensured by the analysis of surface drain
water for oil content. A format could be developed in the same lines as above to
determine losses due to draining. A historical data on oil content will reveal the operation
and maintenance efficiency of the refinery.
Typical values for well maintained seals
100 SCF/ W per rod for reciprocating comp.
8.5 1b/day per seal for centrifugal comp.
4.0 1b/day per seal for pump
These are representative average numbers only and actual rates at a given location for a
given pump or compressor will vary considerably. Replacement of packing by good
quality mechanical seal will reduce the losses. It will also reduce the power consumption
of the drive and hence energy savings. Failure of seals contribute to losses which can be
minimized by prompt rectification / replacement of seals.
Drip Losses:
Drips and leaks add to high losses. A hydrocarbon leak is easily detectable. Necessary
steps should therefore, be taken to stop the same without any loss of time. The table 3.03
shows the amount of leakage loss in litres (approx.)
Table 3.03 . Estimation of Leakage Loss due to drips & leaks
Leakage In one
minute
In one
hour (Lts)
In one
day (Lts)
In one
week
In one
month (Lts)
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
186
(Lts.) (Lts)
1. Drop/sec. 0.003 0.17 4.4 31.0 132
2. Drop/sec. 0.009 0.57 14.6 101.0 398
A drop breaking
into
a stream
0.057 3.89 93.4 634 2705
1/8" Stream 0.65 43.0 1011 6842 29594
3/16" Stream 1.11 70.0 1591 11456 49165
1/4" Stream 2.36 152 3500 248211 106445
Leakage of product not only cost excise duty & product loss but also is a fire hazard.
Leakage in a refinery usually occur at :
Pipe joints
Flanges
Valve glands
Pump glands
Storage tanks
Loading rack
PREVENTION AND REDUCTION OF LEAKAGE LOSSES
a) Leakage through pipe joints
As and when a leakage is detected through a threaded pipe joint, the pipe line should be
immediately repaired. Expansion joint should be provided on long pipeline. Thermal
expansion relief provision must be made. This will take the undue stress and strain of the
pipelines due to variation of day's temperature.
b) Leakage through Flanges
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
187
Flanges should be of standard thickness. The nuts should be tightened crosswise and
diagonally opposite. All nuts should be tightened little by little cross-wise to ensure
equal tightening. Proper gasket packing with liberal use of grease is to be used. Packing
should not extend beyond the flange collar. This helps in tightening the holes evenly all
round.
c) Leakage through valve glands
For globe valves the gland packing should be asbestos rope packing.
For gate valves and valves of bigger diameter where the gap between the valve
stem and the body is more than 3 mm, graphite asbestos rope packing of square
cross section should be used.
The old packing must be removed from the valve and then replace with the new
packing adequately. The gland nuts should be tightened evenly.
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
188
Controlling Steam losses by survey Steam losses occur due to leaks from gaskets, pipe flanges, pin holes, steam trap passing
etc. This may be quantified and exact reason identified for correcting the situation and
minimizing steam leaks. A typical output is given below.
=======================================================
S T E A M L E A K E S T I M A T I O N
ABC Oil Corporation Ltd
=======================================================
User : xxx Refinery
Unit : India
Input file : C:\tthermw\stmleak.in
Output file : C:\tthermw\stmleak.out
Archive ,, : c:\tthermw\stmleak.arc
Date : 01-22-2008\05:54:44
1. P R O G R A M
O U T P U T UNIT TAG LEAKAGE PLUME SIZE STEAM Loss CODE NO CATEGORY (metre) crude 1 a 1.55 44.826 crude 2 b 1.72 61.341 crude 3 a 1.55 44.826 crude 4 b 1.72 61.341 crude 5 a 1.5 40.876 crude 6 b 1.02 16.860 crude 7 a 1.34 30.428 crude2 2 b 0.72 9.693 crude2 1 a 1.25 25.772 crude2 2 b 1.22 24.385 crude2 1 a 1.15 21.430 crude2 2 b 1.02 16.860 crude2 1 a 1.15 21.430 VBU 2 b 1.22 24.385 VBU 1 a 1.15 21.430 VBU 2 b 1.42 35.267
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
189
2.TOTAL STEAM LEAK IDENTIFIED
Total Steam Leaks in kg/h 501.1516
Equivalent Fuel in kg/h 7099.648
Steam Enthalpy for estimation (kcal/kg) : 10200
Fuel Calorific Value in kcal/kg : 720
3. LEAK CATEGORISATION
Code 'a' denotes leaks from pin holes & weld joints
Code 'b' denotes leaks from valve glands.
Code 'c' denotes gasket leak from joints.
Code 'd' denotes gasket miscellaneous leaks
***** end of steam leak program *****
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
190
5
ENCON THRO’ ENERGY EFFICIENT TECHNOLOGY
AND EQUIPMENT
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
191
I. MODERN ENERGY CONSERVATION
1.EMISSIVITY COATING APPLICATIONS
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
192
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
193
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
194
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
195
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
196
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
197
II. OXYGEN ENRICHED AIR :
Air for combustion contains 21% Oxygen by volume. This could be increased by mixing pure
oxygen from other sources ( by product oxygen ) . This reduces the Nitrogen content in the
combustion air and reduces input volume of air. For every 21 nm3/hr of oxygen utilized in any
combustion process, accompanying 79 nm3/hr of nitrogen acts as a dead weight and consumes
energy without doing any work. If there is some source of oxygen available as a byproduct as in
the case of electrolytic process, air separation plants etc, the oxygen concentration in the
combustion air can be increased, which in turn reduces air intake
o Oxygen enriched air increases boiler /efficiency substantially due to lower quantity of dry
gas generation and dry gas losses.
o Besides this, the Flame temperature also increases with O2 enriched air.
o This may be calculated by stoicheometric balance of the system.
Melting and reheating glass and metals are energy intensive processes. Though industrial
furnaces have become much more energy efficient in recent years, oxygen-enriched combustion
technologies can improve their efficiency still more.
o Oxygen enrichment reduces or eliminates the need for combustion air, resulting in less
nitrogen oxide production.
o Oxy-fuel combustion also increases the flame temperature without increasing fuel cost.
Consequently, productivity can be increased while cutting energy use by as much as 50
percent. Oxy-fuel combustion can be implemented as a retrofit or to replace older
technologies.
REPORTED BENEFITS
" Cuts energy consumption 30% to 50%
" Decreases nitrogen oxides up to 90%
" Decreases particulate emissions 30% to 70%
" Eliminates the need for waste heat recuperators
" Increases production 10% to 30%
" Lowers maintenance costs
" Improves temperature stability, heat transfer, and control
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
198
Barriers to Market Acceptance
1. price
2. risk of failure
3. benefits not understood
4. priorities not on benefits of new technology
5. lack of technology awareness
Development Stage
1. need for the technology identified
2. technology concept developed
3. initial research findings reported
4. research on concept completed
5. commercial pilot completed
6. introduction to commercial market
7. immature market demand
8. mature market demand
9. market saturation
. A typical stoicheometric calculation for a boiler plant is as given below.
Item Unit Base case Enriched by 1 %
Carbon Wt % 68.0 68.0
Hydrogen ,, 5.0 5.0
Sulfur ,, 2.0 2.0
Oxygen ,, 0.1 0.1
Ash ,, 24.9 24.9
Theoretical O2 reqd
Wt in kg/per kg fuel 2.211708
Theoretical air reqd ,, 9.4919145 9.161883Actual air @ 20% excess air. .. 11.3902974 10.9942596
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
199
Base case
O2 = ((0.68/12) + 0.05/4 + 0.02/32) – 0.001/32) x 32 = 2.211708 kg / kg fuel
N2 = 7.2802065 kg / kg fuel
Total Theoretical air = 9.4919145 kg / kg fuel
@ 20% excess air Actual air = 11.3902974
1 % Oxygen injection by volume.
Revised air composition : O2 = (22/101) = 21.78 : N2 = 78.22
O2 required = 0.069115875 kg mole / kg fuel
N2 accompanying = 6.950175 kg /kg fuel.
Total enriched air input = 9.161883 kg / kg fuel.
@ 20% excess air actual air = 10.9942596
Reduction in air intake = 3.474341 % (theoretical )
Due to high adiabatic flame temperature achievable by enriched oxygen, the boiler will
show an efficiency improvement between 0.75 to 1%. This will reduce specific fuel consumption
of the boiler further. For a 500 t/hr boiler, the reduction in fuel consumption and air consumption
will be around 1 and 15 t/hr respectively.
TYPICAL PAYBACK
Two years for good applications, based on energy savings. Taking other benefits into account
will shorten the payback period.
WISCONSIN APPLICATIONS
Oxy-fuel combustion is used to melt steel, aluminum, or glass; or to reheat metals and glass for
further processing. In this Oxygen is premixed with the fuel and burnt in the combustion device.
A typical application is the Oxy Acetylene flame.
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
200
Notes :
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
201
III. ON LINE CLEANING
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
202
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
203
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
204
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
205
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
206
IV. High Efficiency Heaters / Boilers
It was shown in previous chapters that all fired equipment such as heater / boiler
operating under design conditions have maximum efficiency. This matches with the design
efficiency.
For increasing the efficiency above the design value, a number of changes in operating
parameters and hard ware configuration may be required. Advanced process control is yet
another concept which is used to enhance the process efficiency.
Advanced Process Control
Models, control and optimization algorithms, and real-time data combine to predict
healthier processes
With advanced process controls (APCs), you either know what they are or you don’t. If
you do know, you know they give for-real, money-in-your-pocket rewards to manufacturers—
rewards so impressive that very few will speak publicly about using them. If you don’t know, it’s
time you did, don’t you think?
As used today, the technology traces its beginnings to the late 1950s. It predicts process
outcomes of varied operating conditions. APC relates manipulated variables (e.g., condition of
a process heater) and control variables (e.g., temperature), providing multivariate control and
also adaptive tuning and predictive / process diagnostics.
Generically, APCs consist of four components:
I. A computer-simulation model that integrates process knowledge
II. Historical data
III. Control and optimization algorithms and
IV. Current, real-time process information.
Justification for implementing advanced process controls comes from improved
performance, because they stabilize operations. They remember different operating scenarios, so
your operators won’t have to. With adaptive control, you could update an existing control model.
For example, if the tool you’re using is a proportional-integral-derivative (PID) tuner, then you
could develop new tuning parameters.
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
207
With predictive control, you handle time delays more effectively in the existing
controller, to optimize throughput and reduce waste.
Real APC performance-killer is inadequate monitoring and maintenance. So before
installing and then ignoring them, consider this: If you use model predictive control (MPC), it
“will degrade 10 percent per year. Typically, at 30 percent degradation, users re-step. The
controllers will either die a smoldering death in three to five years, or very quickly, within six to
12 months, if there’s no proper support.
And support is what manufacturers have less of now, due to downsizing and squeezed
budgets. In general, with fewer control engineers, there’s a pinch on plant performance
monitoring and maintenance. A condition-based approach can overcome this. Process monitoring
tools can also fit into preventive and predictive maintenance strategies. Combining those
technologies with we may get a view into operations via a dashboard. This approach could
reduce maintenance costs by 30 percent and increased profitability by 5 percent.
As to causes of maintenance headaches, about 75 percent of regulatory controls (PID
loops) in a facility are under performing. Many control professionals spend most of their time in
fire-fighting, or reactive, mode. That would mean turning from a reactive / run-to-failure model
and possibly the preventive model, in which assets are often replaced too early, to the forward-
looking predictive-maintenance model.
Better process forecasts come through a breakthrough in process control termed
Predictive control models . In the identification algorithm of the modeling engine component of
its core technology, the software identifies open-loop process models from closed-loop and
dynamic process data. Most users have to break the PID loop and then artificially excite the
process to generate response data to identify the open-loop process models. But the innovative
modeling engine takes process data, without artificial excitation, and extracts the open-loop
process model.
Forward to the future also comes through the other half of his company’s core
technology. It’s a control engine that he says uses the process model selected by the modeling
engines’ identification algorithm “to predict the future direction of the process and generate a
series of control outputs, based on the history of all the variables and predictions.”
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
208
Users find value
And with as many variables encountered in refineries, petrochemicals, food processing,
water treatment and elsewhere, there are many specific ways APCs prove their worth. At Motiva
Enterprises, LLC’s Norco Refining facility near New Orleans, a company-vendor team has
worked for nearly 20 months on installing Shell-licensed-to-Yokogawa multi-variable control
and on-line modeling packages at four process operations, including two ethylene units, says
Merle Likins, a principal advanced control engineer in Yokogawa’s Houston office. The retrofit
involves replacing a pneumatic-instruments system and an older distributed control system
(DCS),
Another process example is at American Water Services Canada Corp., in Grand Bend,
Ontario, Canada. The company uses Mantra advanced controls, from Control Soft Inc., of
Highland Heights, Ohio. This is installed in the pre- and post-chlorination section. This treats
120 million-gallons-per-day (MGD) water-treatment plant in Grand Bend and the 42 MGD Port
Stanley water treatment plant.
The PID loop picks up the pre-chlorination water concentrations. Using the set point that
Mantra automatically controls. it’s flow-paced treatment levels. If flows vary, Mantra changes
the set point, and controls chlorination level.
The controller determines if the chlorine concentration in the end-of-treatment clear wells
is too high or low. If it’s either high or low, then the technology will adjust the level. The
technology, which connects pre- and post-chlorination as well as discharge values, provides an
estimated 25 percent savings , because manual adjustment of chlorination is stopped and the
control is precise with automation.
But regardless of the process in which such controls are used, it is better that the users
monitor the operation. The monitoring must reach into every level of control systems—from the
first-level instrumentation layer, to regulatory controls, then analyzers and online controls and,
finally, APCs themselves.
The issue is not with one or the other layer; it’s with all layers. If any of these break, then
the others can’t deliver. Users must be cognizant of inherent process time issues. In every case in
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
209
a plant where there is a product, there is a time-delay problem. All the subsystems need to be
working perfectly well for the success of the scheme.
Be aware of what causes poor operational performance—controls degrading to the lowest
common denominator—Gough also advises. The result is an insidious ineffectiveness,, because
there’s nothing patently obvious that something is wrong when the user simply detunes the PID
by turning the knob. The symptom goes away.
But in reality, no progress has been made toward finding the solution. From that day
forward, users pay a price in poor performance. Part of the problem, is that working with PID-
type tools means having to do something different each time we work with them.
Also, users need to understand for which plant level the APC’s have been installed.
Historically, that meant something residing at the supervisory level to optimize targets—and
setup has been fairly expensive and complex to implement. Bigger chemicals companies are
finding that traditional APC technologies are improving premium applications, such energy use,
products produced and economic value.
Operational relief also comes because model-predictive technology (MPT) forces the
right answer. The controller needs the right model. That forces the user down a better road. It
forces one to understand what’s going on in the process. So what’s the advanced control
approach? Running different products that cause signal changes in the loop,the controller retrains
under the new conditions, which allows end-users to associate the change with the correlated
process change. And then one can save it (the control change) and recall it.
Users can then build up these sets of models so that, over time, these control systems run
well over all conditions. It evolves to the new set of parameters.
APCs are overcomers and enhancers, with exceptional potential. They leverage existing
knowledge and skill of operators and engineers, real-time process data, dynamic process
variations, advanced mathematics and connectivity. They provide steadiness. They offset or
eliminate poor-performance penalties.
APCs can help manufacturers trade the guess-work and headache involved in PID tuning
for boosted performance. And, trade aggravation for success and higher profitability. In today’s
produce-to-order, zero-defect, just-in-time global market, APCs can keep your company’s clients
and shareholders happy.
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
210
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
211
MULTI FUNCTIONAL ADAPTIVE CONTROL FORFURNACE OPERATION
MFA CONTROL SOLUTION FOR OIL REFINERY FURNACES
Process:
An oil refinery consists of a series of distillation towers and furnaces. Crude oil is piped
through hot furnaces and resulting liquids and vapors are discharged into distillation towers to be
separated into components or fractions by weight and boiling point.
Gasoline, liquid petroleum gas, kerosene, diesel oil, and intermediate streams are
produced.
Goals:
Refinery furnaces consume so much energy that it contributes to a high percentage of
operating costs. It is desirable to tightly control furnace temperatures and other process variables
to optimize separation, minimize energy consumption, and maximize yield.
Challenges:
A typical refinery furnace consists of multiple passes of oil pipes. It is naturally a multi-
variable process with multi-zone temperature control problems. It is difficult to tightly control
the oil temperatures of each pass and outlet due to interactions between the passes and changing
operating conditions.
The distillation tower level and furnace combustion are also critical but difficult to
control.
Solution:
Effective Model-Free Adaptive (MFA) control solutions for controlling critical process
variables without the need to build furnace process models and retune controller parameters.
Tower Level Control:
Use a Robust MFA to smoothly control the distillation tower level and minimize outlet
flow variation to reduce potential vicious cycles in the distillation tower-furnace chain. User-
selectable bounds on level PV protect the level from running too high or too low.
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
212
Furnace Temperatures:
By using a MFA controller to manipulate the oil flow of each pass, interactions between
the temperatures are decoupled. Hence effective temperature control can be achieved. An MFA
controller is able to tightly control the Outlet Temperature.
Anti-delay MFA features may be enabled to handle the large time delays.
Combustion Control:
MFA is also used to control the intake and exhaust fans. MFA can decouple the
interactions of these 2 fans so that the fuel-air ratio can be effectively adjusted to achieve
better combustion efficiency.
Application Story:
Many Refineries have deployed the MFA control system for its vacuum furnace and
distillation tower and the observations were as follows.
Outlet oil temp is controlled within +/- 1 deg C specification;
Temperature deviations between 4 passes are minimized;
Better combustion and level control
Improved production safety
Separation efficiency and
Productivity.
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
213
Use of MFA Control Benefits:
Tightly controls furnace outlet temperature and minimizes deviations of zone
temperatures.
Achieves smoother operations, higher yield, and energy savings.
Decouples loop interactions and minimizes chain reactions among the columns and
furnaces.
Avoids potential vicious cycles, plant up-sets and accidents.
Improves feed throughput and minimizes over/under heating.
Return on investment within a few months.
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
214
Example 2 Heater Pass Flow Control :
Fired heaters used in various process units including the Crude and Vacuum Distillation
units and other processes form the heart of the operation. Besides controlling the operating cost (
energy cost ) , precise feed temperature is extremely important for product separation.
Main objective of the process is to maximize production rate and minimize quality give-
away in various side stream products. This could be achieved only by precise control of feed
temperature and the draw-off temperatures in various sections of the column.
This could be a model based control system in which the flow rate in each coil is
manipulated to get constant pass outlets temperature. Flow deviation from coil to coil is analyzed
and skin temperature of coils present in each pass is compared.
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
215
V. CASE STUDY : BOILER SYSTEM OPTIMIZATION
This study covers the benefits of optimized controls for the Turbo generators TG-II, III,
& IV and their Steam systems only and does not include the newly commissioned TG-V or the
Boilers as the boiler data are very scanty and not sufficient for the benefit study. It has to be
noted here that the benefits realizable will be higher if we include the newly commissioned TG-
V as it will provide more degree of freedom for the control optimizer.
The brief tabulation of the yearly benefit for these 3 TGs is given below. Within the
constraints of utility operational requirements, the power package will calculate the optimum
power targets for these 3 TGs, on a frequent basis ( every 10 minutes once or 30 minutes once ),
which can either be implemented on a closed loop control or displayed as operator advice,
leaving the operator to implement them.
The benefit evaluated for implementing as stated above, is tabulated below for a quick
view of this document.
Table 2.1-1 Benefit summary
Benefits of Optimizing TG-II,III & IV OperationsHP Steam Saving T/Hr 6.01 LP Steam Saving T/Hr 6.98
HP Steam yearly Saving Rs 15151896 HP Steam yearly Saving Rs 8542646
Total yearly Benefit Rs 23694542
The results of the existing operating parameters and their optimum parameters given by
our power package are clearly tabulated and shown on the charts in the following sections of this
document.
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
216
1.0 INTRODUCTION
The unit under study requested Dr.G.G.Rajan, consultant to YIL to study their Utility steam
and Turbo generators and explore the feasibility of implementing Advanced controls in this very
important area of operation at Badrachalm. Subsequently Dr.G.G.Rajan ,Mr.Anil Dutt, and
Mr. Jayabal of YIL visited the site and discussed with the Utility division personnel. The part
of the data gathered then and sent to us by ITC thereafter were the basis of this feasibility report..
The feasibility report in this document is only applicable to TG-II, TG-III & TG-IV and
evolving optimum power targets for them as described below, and the benefits reported in this
document is only with respect to these 3 TGs.
If the new TG-V is included after commissioning it, the degree of freedom for the control
operation will be more and will lead to higher realizable benefits. The optimization of the Boilers
can be taken up later.
2.0 BRIEF DESCRIPTION OF THE TGs & STEAM SYSTEM
ITC had at the time of our visit 4 TGs and the 5th one was being installed. It was explained
to us by the utility staff that TG#1 was not regularly used and only
TG#2,3&4 were used for the generation of the power required, besides the imported
power which is a very small quantity. A brief sketch of the TG & Steam System is givenin the
next sheet.
2.1 Control Requirement Definition of the for the ITC TG-Steam
system
It was understood that the proposed controls will implement new optimum power outputs for the
3 TG s ( TG-II, TG-III & TG-IV) with the following requirements:
TG-IV is on load mode that means the Setpoint of the power MW is changeable by the
Operator.
TG-III is set on frequency mode. TG-IV will be tracking the frequency.
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
217
The implementation of the optimum power targets will be done on a closed loop basis or
open loop advisory displays to the operators so that they can set the targets manually.
Figure 2.1-1 ITC TG & STEAM SYSTEM
3.0 Model for TG power outputs
The model of the TG Power outputs are made using the ITC given data and they are shown
alongwith the actual data in the following charts:
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
218
3.1 TG-II POWER MODEL
Figure 3.1-1 TG-II Power out model
TG II Power Model
2.00
2.50
3.00
3.50
4.00
4.50
5.00
5.50
6.00
6.50
7.00
1 2 3 4 5 6 7 8 9 10 11 12 13 14
Sampling Date
TG-II P
OWER
OUTPU
T MW/H
act model
3.2 TG-III POWER MODEL
Figure 3.2-1 TG-III Power Output Model
TG III Power Model
13.50
14.00
14.50
15.00
15.50
16.00
16.50
1 2 3 4 5 6 7 8 9 10 11 12 13
Sampling Date
TG-III
OUTP
UT PO
WER M
W/H
act model
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
219
3.3 TG-IV POWER MODEL
Figure 3.3-1 TG-IV Power Model
TG IV Power Model
13.00
13.50
14.00
14.50
15.00
15.50
16.00
16.50
17.00
17.50
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
Sampling Date
TG-IV
Powe
r MW
/Hr
actual model
BENEFIT ANALYSIS BASIS
Our power package was used to get the optimum parameters with the constraints requirements of
the ITC TG-STEAM SYSTEM. 14 days of relevant data where all the 3 TGs ( TG-II,III & IV)
are normally operating during the period of 1st July to 31st July 2007 are used in the package to
estimate the parameters such as the steam input, LP / MP extracts, TG-IV Exhaust, TG-II,III &
IV Power outputs with respect to the existing data for that particular samples. The data is shown
in the tabulation as well as charts in the next section for comparison and ITC Operations
assessment. The realizable benefits are calculated from the two steams
Reduction in Total HP Steam in take to TGs.
Reduction in LP Steam venting.
3.4 OBJECTIVE FOR THE CONTROLLER:
Minimize Total HP Steam Intake to TG-II,III &IV, while maintaining stable operation of
the Power generation.
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
220
Minimize LP Steam venting to the extend feasible.
3.5 The constraints which are used are given below
Maintain MP / LP consumption amount as at the beginning of execution of the program
at every cycle.
Maintain total generated power same as at beginning of execution of the program.
Only adjust the power targets of the TGs which are given for the control.
Honor the minimum & maximum setting of the operator for all the targets.
Honor the operator desired Delta changes on these Targets at every execution.
HP to MP & MP to LP PRDS will not be altered and held as they are at the beginning of
execution each time.
The power from the 3 TGs will be adjusted with the above requirements to their
optimum targets with an objective of reducing the HP Steam consumption and also
reduce the LP steam venting.
3.6 COST & ON LINE HOURS
The table given below shows the cost and on stream inputs taken for the benefit estimation: Table 3.6-1 Cost & Basic Assumption
BASIC Assumptions ON Stream Hours per year 8000 Controller ON LINE % 90 HP Steam Cost Rs/ Ton 350 MP Steam Cost Rs/ Ton 200 LP Steam Cost Rs/ Ton 170
3.7 Reduction of Total HP Steam intake to the TGs
The Total HP Steam rates for all the 14 sampling dates are given below: Table 3.7-1 HP Steam Saving
HP Steam Saving SamplingDate OPT_TOT_HP_STM CUR_TOT_HP_STM %Change
1 259.03 267.58 -3.19 2 259.57 267.30 -2.89 3 254.30 255.60 -0.51 4 263.64 270.58 -2.56
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
221
5 255.26 262.79 -2.87 6 244.55 246.54 -0.81 7 262.59 264.96 -0.90 8 263.08 269.33 -2.32 9 251.30 252.96 -0.66 10 247.01 254.96 -3.12 11 259.93 268.63 -3.24 12 241.08 249.25 -3.28 13 250.52 258.30 -3.01 14 250.93 258.17 -2.80
The following chart shows how the existing HP Total steam and Optimum HP Total steam are
varying. You will notice that all the sampling data for 14 days could have been done with less
amount of HP Steam intake versus the existing HP Total steam consumption.
Figure 3.7-1 HP Steam Saving
Total HP STEAM OPTIMUM & Actual
235
240
245
250
255
260
265
270
275
0 2 4 6 8 10 12 14 16
Sample Date data
HP ST
EAM
TOTA
L
OPT_TOT_HP_STMCUR_TOT_HP_STM
3.8 Reduction of LP Vent Steam
The package is programmed to have a stable LP Venting of about 1 to 2 T/Hr, in order to have a
stable Steam header pressures. While reducing the LP Steam venting, it is taken care not to
disturb the LP Steam users by way of maintaining the total LP Steam amount available for the
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
222
LP Steam users. The following table shows how it could have managed against the existing
venting for all the 14 days data.
Table 3.8-1 LP Steam Vent Reduction
LP Steam Vent Reduction T/HR SamplingDate OPT_LP_VENT CUR_LP_VENT %Change
1 1 8 -87.50 2 2 9.9 -79.80 3 2 7.3 -72.60 4 2 7.08 -71.75 5 2 6.79 -70.54 6 2 6.71 -70.19 7 2 3.13 -36.10 8 2 5.42 -63.10 9 2 7.25 -72.41
10 2 5.58 -64.16 11 2 8.63 -76.83 12 2 20.79 -90.38 13 2 12.67 -84.21 14 2 15.46 -87.06
Figure 3.8-1 LP Steam venting OPTIMUM Vs the existing LP Vent
LP VENT CURRENT Vs OPTIMUM
0
5
10
15
20
25
0 2 4 6 8 10 12 14 16
Smpling Date
LP V
ENT
T/HR
OPT_LP_VENTCUR_LP_VENT
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
223
3.9 Estimated yearly benefits for the Optimum Control
HP Yearly benefit = Average Hourly HP Steam saving * 8000 * 0.9 * 350
LP Steam venting reduction yearly benefit = Average Hourly venting reduction * 8000 *
0.9 * 170
The estimated benefit is given in the following table:
Table 3.9-1 Estimated yearly Benefits
Benefits of Optimizing TG-II,III & IV OperationsHP Steam Saving T/Hr 6.01 LP Steam Saving T/Hr 6.98
HP Steam yearly Saving Rs 15151896 HP Steam yearly Saving Rs 8542646
Total yearly Benefit Rs 23694542
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
224
VI. MONITORING BOILER EFFICIENCY BY SIMULATION MODELS.
Figure given below represents an EORT Simulation program data input.
Parameters used in the simulation model are
C/H ratio of fuel
HHV in Kcal / Kg
Air temperature
Humidity
Blow down rate
Flue gas temperature
O2 in flue gas
% Conv & Radiation loss
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
225
The simulation will predict the boiler efficiency at the observed conditions. This can be used
as an excellent tool for boiler performance monitoring. The output after data entry is as shown
below.
Under these operating parameters, the simulated boiler efficiency will be 81.52 %. If the
observed efficiency is lower, the reason may be investigated by detailed analysis.
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
226
VII. COOLING WATER SYSTEM
Cooling water is one of the major utilities in power industries. It is a very important
utility from cooling and utility cost point of view. Water is also used for drinking , generation of
steam by treating the fresh water to Boiler Feed Water. Water treatment cost increases the
utility cost and ineffective treatments lead to a number of problems.
Monitoring of cooling water quality cannot be under estimated as cooling water quality is
responsible for scale formation and loss of heat transfer efficiency , corrosion of pipe lines and
process equipment failures.
Cooling water is extensively used in surface condensers, inter-stage coolers, vapor
condensers etc. Fouling of these condensers could be due to organic or inorganic matter present
in the water in the form of bacteria and/or dissolved salts . The quality of water is maintained
by water treatment which involves addition of appropriate chemicals in calculated rates.
Depending on the dissolved salts and gases , the water may be either scale forming or corrosive
and the water must be in a balanced state( neither scale forming nor corrosive).
Cooling Water flows in a main header , sizes ranging from 600 to 1500 mm diameter
from where the individual units draw water in required quantities and the outlet from units join a
return header, which enters the top of the cooling tower and is cooled by counter current air
flow. The entire flow circuit is prone to fouling or corrosion and depends on the location and the
process. The magnitude of fouling/corrosion problem depends on a number of process
parameters.
Water Management:The objective of water management is to
* minimise water consumption
* reduce water wastage
* eliminate scale formation and corrosion
* minimise water treatment and utility cost
* identify system condition and approriate treatment programs.
All these activities have cost implications and hence monitoring the actual performance of
the complicated system warrants using computer-based models.
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
227
a.Specific Consumption Model: Specific Consumption of cooling water controls the operating cost by reducing
I. Electric power in cooling water pumps
II. Fin-fan load
III. Cooling water treatment cost and
IV. Blow down rates.
This cost may be dominant in water-scarce areas.
A typical specific consumption model for cooling water consumption in a cryogenic plant
is given below. Similar models could be developed for all the process units and compared with
actual consumption for remedial action. Data input for the model are Capacity utilization of
plant, cooling water circulated and the running hours. Specific consumption of cooling water is
defined as the water circulated in m3 per ton of feed.
The model is of the form
cw = a * cu ^3 + b*cu^2+c*cu+d
where
cw is specific consumption of water in m3/t feed
cu is capacity utilization or load factor of the unit in %
a,b,c&d are constants.
For the same capacity utilization and process conditions, specific consumption of cooling
water and hence the utility increase with passage of time due to fouling and deterioration of
equipment efficiency Abnormal specific consumption warrants a thorough process
investigation.
b. Trouble shooting :
Computer-based models may be used for trouble shooting cooling water problems
associated with corrosion and scaling. Conventional techniques of monitoring Scaling /
corrosion tendency include
I. Pressure drop measurements
II. Corrosion coupons
III. Iron count monitoring etc.
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
228
When corrosion rates are observed high, the water quality may be observed to be in the
corrosive range and vice versa. Hence for effective corrosion/scale control, the water quality is
monitored by associated parameters like TDS, Calcium Hardness, P alkalinity, M alkalinity,
water pH, temperature etc.
A simple multi -variable model that could be used for identifying corrosivity or scale
forming tendency is given by
LI= a + (-0.01844* t + 2.5058) + (1.006578 log CH - 0.4081191) + (1.006578 log M -
0.0081191)
where
LI = Langlier Index
t = Cooling Water temperature in oC
CH = Calcium Hardness of water in ppm
M=Methyl Orange alkalinity in ppm
a = constant ( 0.1 for tds < 300 & 0.2 for tds > 300)
Using the above model, LI has been arrived at and the value of LI is used to estimate
corrosion or scale forming tendency by a polynomial model as shown below.
corrosion mpy = - 0.1504*(LI)3 - 0.5160* (LI)2 -0.9674*LI + 2.9849
Standard Error of the model : 0.5270
If LI is negative, the system has scale forming tendency and if LI is positive, corrosion is
dominant as given below.
Situation 1 and 2 indicate scale formation while situation 4 shows extreme corrosion rate
follwed by 3 & 5.
Lang Z1 Simulated
index corrosion
2.0 2.0000 -2.2172
2.5 2.5000 -5.0088
-2.9 -2.9000 5.1189
-5.0 -5.0000 13.7219
0.0 0.0000 2.9849
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
229
This is applicable only for inorganic fouling and corrosion. If the observed conditions
deviate from the simulated model , this could be due to bio-fouling and/or biological corrosion,
on account of iron reducing / sulphate reducing bacteria .
Even this type of situation could be estimated by using the cooling water related
parameters like BOD (Biological Oxygen Demand) , loss on ignition , chlorine demand etc .
Conclusions:
Utilities Management include compressed air , refrigeration , fuel etc . The same
modeling methodology could be used effectively for developing performance models for the
total system as explained above. Basic advantages of using the model is , it simplifies a
complex problem into quantifiable data for interpretation. Hence action can be taken in the
right direction after checking the model validity.
Since these models are based on actual operating data and are developed by technical
personnel with 'hands-on' experience , they may be used for training , improve the process
design and for product development .
However variables used in these EORT / Statistical Models must be chosen with great
care to avoid duplication or strongly related variables.
When once developed and implemented , this could save considerable time and money
and other valuable resources.
Utility Management models improve the productivity of the overall system , offer
better operation , long run lengths , minimum failures, planned maintenance, minimum
operating cost etc.
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
230
VIII. Performance Monitoring of steam turbines
While turbine efficiency is calculated by energy input / output analysis, performance of steam
turbines vary with the load factor, aging /on -stream hours /days , steam quality etc. Turbine
manufacturers often supply the performance characteristics of each machine in the form of a
curve for various energy input conditions.
Actual performance of the turbine is compared to the base case to determine whether or not
the efficiency is within the acceptable limits. A typical steam turbine characteristic is given in
fig 5.1(a).
This characteristic curve gives the input steam rate vs output in kw for any chosen
extraction steam rate. The graph is drawn from the actual performance data for a machine. For
performance monitoring of the turbine , the output observed is compared against the base
case ( in this case characteristics curve) and checked whether the performance is normal or
not. Case given in fig 5(a) has extraction rates of 0 ,5 ,10 and 15 % of steam input and any
intermediate value is interpolated. An example of how the turbine characteristics may be used
for performance monitoring is given in the section.
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
231
observed data:
1. Steam inlet 150.0 t/h
2. Pass - out 0.0 t/h
3.Power output 20 mw
From the characteristics, power output should be 22.75mw.
Observed Deviation in performance is -2.75 mw
% deviation from base value = (-2.75 /22.75)* 100
= 12.08 %
The deviation is high and needs a thorough investigation. If the turbine operation continues
under this condition, excess steam consumption would have been about 15 t/hr for the same
output. ( given by y-x )
Increase in annual steam consumption based on 8000 hrs of operation works out to 120000
tons. For applying this monitoring technique, user must have the complete turbine characteristic
for each machine and refer to the same every time the checking is required.
Computer-aided performance monitoring involves mainly the collection of performance data
for the turbines operating under varying loads and other parameters and using these data to
develop time-dependent and non-linear / linear multi variable models for performance
prediction.
Computer-aided performance monitoring of steam turbine involves .
1.Turbine data collection. ( energy input / output information )
2.Evaluation of enthalpies of various streams
3.Developing material balance
4.Developing energy balance.
5.Efficiency calculation for the observed conditions.
6.Deviation Analysis with base data.
7.Identifying causes for corrective action.
Performance of turbines vary with time and connected load as shown in the above table.
As the shaft load increases , efficiency of the turbine increases and the specific consumption of
energy decreases . It is possible to develop a suitable performance model based on the actual
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
232
observations. Necessary flow rate corrections are to be applied for the steam pressure,
temperature and flow rates.
Typical output of steam turbine models using the above data is given in box no 5.1. Box no
5.2 gives the simulated turbine efficiency for new conditions, based on the performance model
developed by this method.
Box 5.1 Steam Turbine Performance Model (load % vs efficiency)
Box 5.2 Simulated Performance for new conditions.
load% X1 efficiency % observed simulated
25 25.0000 21.1000 21.1000
50 50.0000 30.7000 30.7001
75 75.0000 38.2000 38.1999
100 100.0000 40.0000 40.0000
S.E of Model : 0.0001
load% Z1 simulated efficiency
20 20.0000 19.2446
35 35.0000 24.9617
65 65.0000 35.6824
85 85.0000 39.8055
105 105.0000 39.3594
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
233
Steam Turbine Performance Model 2 ( load vs sp.cons)
load X1 observed simulated
sp.con sp.con
25 25.0000 21.6100 21.6100
50 50.0000 15.5200 15.5199
75 75.0000 12.2100 12.2101
100 100.0000 11.3900 11.3900
S I M U L A T E D O U T P U T F O R N E W D A T A
load Z1 simulated load Z1 simulated sp.cons sp.cons
20 20.0000 23.1873 65 65.0000 13.2190
35 35.0000 18.8217 85 85.0000 11.5996
When once the models are developed , it is possible to determine the performance level of
each machine with reasonable accuracy at any point of time for various connected load
conditions. In the case of gas turbines , the above models may be modified to give specific
energy input / kw of net shaft power and efficiency vs load. This technique is superior to the
conventional static performance monitoring method, which does not account for random /
variable parameters.
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
234
IX. CENTRIFUGAL FANS / BLOWERS
The operating principle of centrifugal fans is exactly the same as that of centrifugal pumps. Gas
enters at the axis of the impeller and is thrown outward by the vanes into a scroll. The clearances
are large, and the discharge heads are low. Because of the low density of the gas, fans rarely
discharge at more than 60 in. H20, and often at 5 to 10 in H2O. Sometimes, as in ventilating
fans, all the added energy is converted to velocity energy and almost none to static head.
In any case the gain in velocity absorbs an appreciable fraction of the added energy., and must be
considered in estimating efficiency and power requirements. The compressibility of the gas,
however, may be neglected.
The static efficiency of a fan is the fraction of the shaft-work input to the fan that is converted to
pressure energy. It is of interest when the fan is used primarily to increase pressure.
The total efficiency is the fraction of the shaft work appearing as both pressure and The
dynamic efficiency is the fraction of the shaft work converted to velocity energy. It is important
when the fan is used to increase velocity rather than pressure and velocity energy.
The power required by a fan is obtained from Eq. (1-1), with the aid of ,the appropriate fan
efficiency.
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
235
For example, if the static efficiency is used, the velocity in the exit gas is neglected, and the
developed head is, by Eq.(1-2)
If q is the volumetric rate of flow at density , in cubic feet per minute, and if p' is the pressure
increase in pounds force per square inch, q = w / , and
Also, when the pressure increase is negligible, the developed head is given by
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
236
Note that at higher RPM, pressure developed is higher for the same air flow rate. Figure given
below is the characteristic curve of the MAB. Power consumption of the blower for a particular
air flow rate and RPM is also given in the table below.
Energy Efficiency Optimization Training Program – Dr.G.G.Rajan
237
Fig 2. Air Flow rate vs Power Consumed (MAB)
Energy in the Global Value Chain –
9-December ’11
Speaker: Dr.G.G.Rajan / Cochin / India
UNDERSTANDING ENERGY EFFICIENCY OPTIMIZATION
12/12/2011
Energy Efficiency Optimization
• Energy efficiency optimization refers to the application of mathematical / statistical / operations research techniques to
• Minimize energy consumption• Energy cost and • Loss reduction without loss of
production quantity / quality.
Tools for EE optimization
Desirable : Basic mathematical and / or Operations Research background. Theoretical concepts of optimization.Energy efficiency optimization tools such as Optimization and modeling software
Boiler Example – Load optimizationThree boilers B1,B2 & B3 are in operation in a process unit, generating 100, 150 and 180 t/hr of steam respectively.
Steam / Fuel ratio which is function of capacity utilization of each boiler is given below.
The design capacity of the boiler B1,B2 & B3 are 120, 180 & 210 tons/hr respectively.
Determine the optimal load, the boilers should handle to meet the steam demand of 430 t/hr at minimum fuel consumption.
load vs boiler efficiency
11.5
12.0
12.5
13.0
13.5
14.0
14.5
15.0
15.5
16.0
16.5
60% 70% 80% 90% 100%
load
effi
%
boiler1 boiler2 boiler3
Table shows the information in terms of steam produced in tons/hr versus fuel in kg, which will be used in boiler load optimization.
From the above data regression equations have been developed using MS Excel program.
y1,y2 and y3 are quantity of fuel consumed for x1,x2 & x3 t/hr of steam generated in boilers 1,2 & 3.
Total quantity of fuel consumed = y1 + y2 + y3 , which has to beminimized.
Regression equations for fuel consumption as a function of Boiler loads for boilers 1,2 and 3 ( using Excel Spreadsheet ) are
y1 = 53.52381* x1 + 2196.9
y2=56.911678*x2 + 2124.82
y3 = 47.496883*x3+3191.98
Total fuel consumed at steam production rates x1,x2 & x3 are
F = 53.52381* x1+56.911678*x2+47.496883*x3+ 7513.7
The ultimate linear programming model for minimizing fuel consumption is given by
Minimize 53.52381* x1+56.911678*x2+47.496883*x3+ 7513.7Subject to x1+x2+x3 =430 ( steam demand )x1<=120 ( boiler 1 capacity )x2 <=180 ( boiler 2 capacity )x3<= 210 ( boiler 3 capacity )x1>=0x2>=0x3>=0
Solution (From Lingo Program)LP OPTIMUM FOUND AT STEP 2
OBJECTIVE FUNCTION VALUE = 22088.37 kg fuel
VARIABLE VALUE REDUCED COST
X1 120.000000 0.000000
X2 100.000000 0.000000
X3 210.000000 0.000000
NO. ITERATIONS = 2
As per existing operating pattern presented above, the fuel consumption is calculated as 29952.3 kg/hr which is higher than the optimum value by 7863.9 kg/hr. i.e. 35.6 % than the optimum consumption.
This example shows how operation of boilers can be optimized to minimize fuel consumption using operations research techniques.
Existing load vs Optimal load
22088.429952.3Fuelconsumed
430430Total
210180Boiler B3
100150Boiler B2
120100Boiler B1
Optimalload t/hr
Existingload t/hr
Boiler name
Optimization stepsThis is referred to as ‘Problem Formulation’.
This is the most critical aspect of any optimization problem to achieve tangible results.
Any snag in the problem formulation, may result in non-feasible / impractical solutions.
Extreme care must be taken in ‘problem formulation’, as this is the most intelligent activity to tackle any problem.
Following steps are involved in optimizing the performance of any system.
First step is to define the objective / goal we are aiming at.
Examples given below are typical objective functions related to production or operation of an industry.
They are
� Maximizing production quantity
� Maximizing operating profits
� Minimizing operating costs
� Minimizing energy consumption / energy costs
� Minimizing emissions level
� Loss reduction etc
Step 2: Identify process constraints which have an impact on theobjective function. Typical examples are
Capacity utilization or load factor in boilers, heaters, fans, blowers, turbines etc.
(As capacity utilization increases, specific energy consumption reduces and efficiency increases up to certain capacity (i.e. design capacity ) and starts dropping down, when the capacity utilization increases further.)
Run length .(In many equipments, wear and tear results in loss of efficiency which will result in higher energy consumption.
In the case of boilers, heaters, heat exchangers etc fouling increases with operating period, which retard heat transfer and decrease equipment efficiency and increase fuel consumption )
Operating severity. This refers to the reaction temperature / pressure / recycle etc which reduce the energy efficiency of thesystem. In the case of conversion processes, the objective is toincrease conversion at the expense of energy / operating costs.
Feed quality / composition : In petroleum refining, petrochemical, fertilizer plants etc feed quality / composition plays an important part in operating profit, production rate etc.
Step 3: Establish a mathematical model between these variables and the objective function. If these variables are linearly connected, the problem may be solved by Linear Programming methods else it has to be solved by NLP / Parametric programming methods.
In the boiler loading example , relationship between fuel consumption and steam generation quantity was established using linear regression models for all the three boilers.
Using these models, total fuel consumption was established by summing up these relationship to get an objective function equation.
In the case study, where five boilers were considered, a non linear model was developed to determine overall boiler efficiency by individual efficiency models. In this case the objective function was to maximize overall efficiency of the system of five operating boilers.
Step 4: Identify the constraints which impede / control the objective function. In the example dealing with three boilers, the constraint was basically the maximum steam generating capacity. There could be any number of constraints which will be incorporated in the LP model.
Step 5: When once all the above steps have been applied and problem formulated successfully, the solution may be arrived at using Operations Research tools such as Linear / Non Linear / Parametric programming methods.
Logical programmingThis is an optimization technique using logical data analysis.
Example : A power generation unit has five nos of power boilers B1,B2,B3,B4 & B5 whose operating parameters are given in the following table.
In this case boiler efficiencies are linear and proportional to actual steam production.
Hence, it is logical to load the boiler which gives maximum efficiency.
In this case boiler B5 shows an efficiency of 92.5 % at full load of 400 t/hr.
Hence B5 load is fixed at 400 t/hr.
Balance steam to be generated out of other boilers is 900 ton/hr and boilers available are B1,B2,B3 and B4.
Out of these four boilers, B4 shows an efficiency of 91 % at full load.
Hence B4 load is taken as 350 t/hr.
Other boilers left out are B1,B2 and B3 and the balance steam demand is 550 tons/hr.
Next boiler in the efficiency hierarchy is B1 which may be loaded to 300 t/hr.
Balance demand of 250 t/hr may be met by boiler B2.
B3 will be just idling.
This is the simplest logical method followed by many plant managers in real life situations.
In this case, the overall efficiency of the boiler system shall be
Effo = (400 x 0.925 + 350 x 0.91 + 300 x 0.875 + 250 x 0.85) / 1300
= 89.5
Fuel consumption = (400/12.5) + (350/12.0)+ (300/11.5) + (250/11.0)
= 32 + 29.17 + 26.09 + 22.73 = 110.53
Average steam / fuel = 11.76
When the steam demand changes, this exercise has to be repeated time and again .
Manual calculation being a cumbersome process, this may be programmed using operations research techniques as shown in example 1 .
WHY ENERGY EFFICIENCY OPTIMIZATION ?
Energy efficiency optimization is the best route to meet the energy demand at minimum cost without loss of production / output of the system.
Conventional Energy demand management, isguided by the intuitive decisions of the operator, which may not be optimal in most cases.
For effective energy demand management at minimum cost and meeting all the imposed constraints, EEO is the apt solution.
What is optimization ?All optimization problems are made up of three basic ingredients:
An objective function which we want to minimize or maximize.
For instance, in a chemical industry, we might want to maximize the production or minimize the operating cost.
In fitting experimental data to a user-defined model, we might minimize the total deviation of observed data from predictions based on the model.
In designing a process equipment , we might want to maximize the energy efficiency.
Note : the objective function should be quantitative/
Contd …
We should identify the set of unknowns or variables which affect the value of the objective function.
In the manufacturing problem, the variables might include the amounts of different resources used or the time spent on each activity.
In fitting-the-data problem, the unknowns are the parameters that define the model.
In the panel design problem, the variables used define the shape and dimensions of the panel.
Contd…A set of constraints that allow the unknowns to take on certain values but exclude others.
For the manufacturing problem, it does not make sense to spend a negative amount of time on any activity, so we constrain all the "time" variables to be non-negative.
In the equipment design problem, we would probably want to limit the size of the product to constrain its plinth area requirement .
• The optimization problem: • Finds values of the variables that minimize
or maximize the objective function while satisfying the constraints.
• Identifies whether all these ingredients are necessary
• Objective function :• Almost all optimization problems have a
single objective function.
• The two interesting exceptions are• No objective function. • In some cases (for example, design of
integrated circuit layouts), the goal is to find a set of variables that satisfies the constraints of the model.
• The user does not particularly want to optimize anything so there is no reason to define an objective function.
• This type of problems is usually called a feasibility problem.
• Variables • These are essential. • If there are no variables, we
cannot define the objective function and the problem constraints.
• Constraints • Constraints are not essential. • In real life problems, constraints are a
reality, which must be considered in the formulation.
• In fact, the field of unconstrained optimization is a large and important one for which a lot of algorithms and software are available.
• It's been argued that almost all problems really do have constraints.
• In practice constraints are encountered in day to day operation of the enterprise and these constraints will have to be necessarily considered for achieving maximum profits and productivity.
• It is applicable for all enterprises in the economic scenario.
Energy efficiency optimization Macro level :
In big corporations / enterprises having a number of process / power plants (e.g: National thermal power plants), energy efficiency optimization of the total organization will bring down the operating cost, as this approach increases energy efficiency, reduces fuel consumption / cost and pollution levels.
National level :
At national level, energy resource mix may be optimized to meet the energy demand at minimum cost. This concept can improve national productivity substantially.
Unit level Energy optimization At unit level energy efficiency optimization, total unit may be divided into systems,subsystems and equipments.
Their energy consumption / generation data is collected and evaluated.
Taking the actual constraints imposed, an optimization model is developed with the objective of minimizing energy consumption and at the same time without loss of production.
In modern optimization models, Sulfurous emissions are also incorporated in the model to optimize the energy resource mix that will meet all the requisites.
Since this is dynamic, the evaluation must also be carried out more frequently.
Need for Energy Efficiency Optimization
High energy cost produced from primary and secondary sources, warrant maximization of energy efficiency in production as well as consumption.
In operation of boilers, heaters, pumps, compressors, turbines etc % Load on the equipment has an impact on the energy efficiency of the system.
A typical load vs efficiency is shown in the next slide.
In these equipments, efficiency increases with load, reaches a maximum and then starts dropping down.
When a number of such equipments is in operation, as in boilers, an optimum operation of boilers must be chosen to minimize energy consumption and operating cost.
Single variable optimization model
This model is of the form
eff = a * x 2 + b * x + c
where a,b and c are constants and
x is the % load on the equipment in operation and
eff is the % efficiency of the equipment.
The model is developed using the plant data or test runs in which the load is varied keeping all the other parameters constant.
The observed efficiency is calculated and the data entered in the model.
Example
80.012083.511087.010082.59076.08070.07060.060
Equipment Efficiency %
Load %
Observed Data vs Model
60.0
65.0
70.0
75.0
80.0
85.0
90.0
60 70 80 90 100 110 120
load %
effici
ency
%
actual model
Using models in optimization of total system
These simple models are very useful for energy efficiency optimization of the total system.
Consider 4 boilers B1,B2,B3 & B4
Load on boilers x1,x2,x3 & x4 % on design.
Eff % on each model is of the form e1= a1*x12 + b1 * x1 + c1
Similar models are generated for other boilers.
Fuel consumption for each boiler = f1 =
(x1/100)*d1*enthalpy *(1/cv)* (100/e1)
cv = fuel calorific value in kcal/kg, enthalpy of steam = kcal/kg steam produced, d1= design capacity in kg/hr, e1 = eff % from model
Typical total optimization model
Minimize f1+ f2 + f3 + f4
subject to
x1+x2+x3+x4 = demand
x1 >=0.6 * d1
x2 >=1.0 * d2
x2 <= 1.105 * d2
etc
Two variable modelIn this case two opposing parameters have an impact on efficiency, energy loss or operating cost. The objective is to optimize the common parameter that will minimize the operating cost.
Typical example is the insulation thickness.
As the insulation thickness for the same service increases, the fixed cost component increases.
At the same time, energy loss reduces and cost of energy loss reduces.
Total cost = (cost of insulation + energy loss cost ) annualized
The data can be converted into models and optimization carried out.
Insulation thickness optimization model
For developing this model, certain calculations will have to be made.
Total insulation cost for various thickness of insulation.
Annualization of the fixed cost taking into consideration the life of insulation and interest on capital.
This cost varies with
Type of insulation material
Life of insulation material
Operation & maintenance costs
Insulation efficiency
Cost of energy loss for various insulation thickness. This may be calculated by energy loss divided by cost of energy in US$ / Pounds etc per year.
The loss reduces with insulation thickness.
Cost of energy loss is dynamic and varies with the fuel cost. Hence it is imperative to carry out optimization of the insulation systems more frequently, especially when the energy cost goes up disproportionately.
Next slide shows the impact of insulation thickness on operating costs of the system.
Impact of insulation thickness on operating costs
0
20
40
60
80
100
120
140
160
10 20 30 40 50 60 70 80
insulation thickness mm
opera
ting c
ost '0
00 us
$/yr
fixed cost nrg loss total cost
• Pl note fixed cost of insulation increases with thickness and cost of energy loss reduces.
• Optimum thickness in this case is 40 mm at which the total cost is 110 thousand us$/yr
• This varies from system to system
Mathematical model
• This may be converted into a mathematical model by using regression methods.
• These models are • Insulation cost model• Energy loss cost model and• Total cost model• Data used in the program is given in
the next slide.
Data used in the model
134
128
120
114
108
116.5
123
130
Totalcost
567880
606870
625860
654950
703840
88.52830
1051820
1201010
Cost of energy loss
Cost of insulation
Insulationthickness mm
Methodology• There are two methods by which the optimum
insulation thickness may be arrived.• In the first method, two models are developed
for the insulation cost and energy loss cost. The first model is linear and the second model is non-linear.
• These two models are integrated to form the third and final model.
• This model is used for determining the optimum insulation thickness. ( Use regression equations to develop the models )
ModelsInsulation cost model :
This is of the form C1 = A*x + B
Cost of Energy loss:
This is of the form C2 = A2*x2 + B2 * x + C2
Total Cost model:
This is equal to :
C1+C2 = C = A2*x2 + x * ( A + B2) + ( B + C2)
Where A,A2,B2,C & C2 are constants and x is the insulation thickness in mm.
Models
0
20
40
60
80
100
120
140
160
10 20 30 40 50 60 70 80
insulation thickness in mm
'000
US$
/ yr
insuln cost cost of nrg loss total cost
Models for insulation systemInsulation cost model
C1 = .9845237 * x - .9285644 ( S.E. 0.5989)
Cost of Energy Loss model
C2 = 1.610116E-02 * x2 - 2.350891* x + 143.0446
(S.E. : 2.4601)
Total cost model Ct = C1 + C2
= 1.610116E-02 * x2 – 1.366673 * x + 142.1160356For determining minimum total cost, total cost function is differentiated WRT x and equated to 0.
When second order differential is –ve, then x represents the thickness at which the total cost is minimum.
dct/dx = 2 * (1.610116E-02 ) * x – 1.366673 = 0
i.e. 2 * 0.01610116 * x = 1.366673
i.e. x = 1.366673 / (2 * 0.01610116) = 42.44 mm
d2ct/dx2 = 0.03220232 ( + ve)
Hence ct is minimum at x = 42.44 mm
Total cost ct = 1.610116E-02 * (42.44)2 – 1.366673 * (42.44)
+ 142.1160356
= 113.115 ‘000 US$
Refer to earlier slide and note the minimum point is between 40 to 45 mm thickness.
Time dependant model• While the model given in this section refers
to a totally new scheme, there is a need to change / replace insulation after certain period of time, because of the deterioration of insulation material and increase in heat loss.
• Next figure shows the optimum replacement time for the insulation, based on the energy loss data.
02468
1012141618
Rs '0
0000
0
1 2 3 4year
heat loss opr.cost savings
When the number of variables are more, the objective function will be generated using the operating data and converting them into appropriate models.
Then these models may be used in LP or NLP programs to optimize the objective function, within the stipulated constraints.
Refer to the book ‘Practical Energy Efficiency Optimization’ by Dr.G.G.Rajanfor more details
Typical optimization problem
4 7 0 .0
4 8 0 .0
4 9 0 .0
5 0 0 .0
5 1 0 .0
5 2 0 .0
5 3 0 .0
5 4 0 .0
5 5 0 .0
t'pu t 0 0 0 to n s
sp.e
nerg
y '0
00 k
cal/t
s p.e n e rg y
s p.e n e rg y 5 4 0 .7 5 2 9 .9 4 7 8 .7 5 2 4 .5 5 2 4 .0
5 9 8 6 7 5 8 7 0 9 6 0 1 0 0 0
Specific total cost vs thro’put
6 7
6 8
6 9
7 0
7 1
7 2
7 3
t' pu t ' 0 0 0 M T
sp.to
tal c
ost u
s$
s p.to ta l c o s t
s p.to ta l c o s t 6 7 .8 7 6 7 .0 6 6 7 .0 7 6 7 .9 1 6 9 .5 7 7 2 .0 7
7 5 0 8 0 0 8 5 0 9 0 0 9 5 0 1 0 0 0
Total optimization – case studyIn total optimization study, the objective is to maximize profitability of the process of which energy consumption is one of the parameters.
In the case of thermal cracking operation for example,•High severity of operation increases conversion and product yield.•High severity increases energy consumption.•High severity increases operation and maintenance cost.•High severity reduces run length / production per cycle.•Increases failure rate / replacement cost. •The objective here is to determine the operation severity that will minimize the total cost or maximize operating profit.
ConstrainsOperational constrains in this case are
•Coil outlet temperature of the heater ( < 560 oC )
•Coil wall temperature ( < 750 oC )
•Heat transfer limitations in condensers / coolers
•Pressure drop
•Main column performance / product pattern
•Gas production / handling limitations etc
12/12/2011
Advantages
• Use of energy efficiency optimization on continuous basis pays off in a few days to weeks.
• Identifies equipment deterioration.• Longer run lengths / higher production /
power generation of units could be achieved.
• Energy costs could be minimized • Net profit could be increased.
Example - Equipment Maintenance / Replacement Decisions.
• These models may be used for taking equipmentmaintenance / replacement decisions to minimize the total cost of operation which determines the profitability and productivity of the industry.
020406080
100120140
I II III IV V
oprtg fixd total
Economics of Insulation
02468
1012141618
Rs '000000
1 2 3 4year
heat loss opr.cost savings
• This approach evaluates heat loss from various sections of the equipment, which could be used to identify the source of loss for taking insulation retrofit decisions.
12/12/2011
Energy Efficiency Optimization in Typon
Power Plant
A proposal
12/12/2011
ObservationsTypon Power Generation Plant has ten sets of
Boilers and Steam turbines to generate power of the order of 800 MW.
Except a stand by unit, all the boilers and turbines operate at 100 % of the capacity and many even beyond 100 % in view of the power demand.
Some of the boilers indicate an efficiency of 81 to 82% ,while others show an efficiency of 85%.
The overall efficiency of the plant is between 82 to 83.
A systematic energy efficiency study of the total system may boost the overall efficiency to 84 to 85 or even more.
12/12/2011
ObservationsIndividual boiler efficiency is calculated by indirect method ( heat loss)
Direct method is not feasible due to lack of facility for metering coal flow to each boiler.
Indirect boiler efficiency calculation is based on ASME PTC 4.1
For this method, the ultimate analysis of coal used in each boiler must be known precisely. This will be used to determine the dry gas loss and other losses.
Impact of coal quality on boiler efficiency is shown in the next slide.
12/12/2011
ASH CONTENT IN COAL vs EFFICIENCY (LHV)
82
83
84
85
86
87
88
'16 % ' '21 %' '30 %' '35 %'
% eff blr
12/12/2011
PARAMETERS USED IN EFFICIENCY CALCULATION
35.030.016.0 21.0Ash
2.22.22.2 2.2Nitrogen
0.40.40.4 0.4Sulfur19.419.419.4 19.4Oxygen
00.000.000.0 00.0Moisture
3.03.03.0 3.0Hydrogen
40.045.059.0 54.0CarbonWt %Wt %Wt % Wt %Component
12/12/2011
MethodologyWe propose to adopt the following methodology
to improve the overall energy efficiency of the system.
1. Study of coal fuel system covering coal composition to coal particle size. At present 75 to 85 % passes thro’ 200 micron. This may further be optimized .
2. This calls for a process test run , data collection and modeling.
3. Next study will focus on optimal fuel mix for the boilers. ( may require some revamping)
12/12/2011
Contd ….
4. Excess air optimization
5. Though it is felt that lower excess air will improve boiler efficiency, this needs an optimization study, as the flue gas rate is linked to excess air, which in turn affects the performance of air pre heaters.
6. Air pre-heater performance evaluation and optimization covering heat transfer efficiency, leakage rate and other problems identified.
7. Economizer performance optimization.Basically heat transfer efficiency, fouling rate estimation etc
8. Steam super-heater performance optimization
12/12/2011
Contd …..
9. Optimal super heated steam temperature.
10. A revision in the super heated steam temperature by 5 to 6 oC is envisaged. As against the design value of 510 oC, we may maintain 515 oC. A test run is required to firm up this decision.
11. Turbine efficiency optimization study from charcteristics & test runs
12. Optimizing BFW / ID Fan / FD fan drives.
13. Surface condenser performance evaluation
14. CT fans / CW pumps evaluation.
12/12/2011
Optimization study benefits1. It is envisaged that optimization study of the
above listed equipments will improve the overall efficiency of the plant further.
2. Load optimization on boilers and turbines can improve the system efficiency further.
3. Study of other equipments, will boost the performance further and probably reveal scope for cogeneration / tri-generation projects in the long run.
4. Even retrofitting the existing boiler / turbine capacities may be explored.
12/12/2011
Capacity Utilization vsEfficiency
Capacity utilization plays an important role on system efficiency.In normal case, higher the capacity utilization, higher is the efficiency. This is known as system characteristics. Operational capacity must be optimized to achieve highest possible efficiency of the total system to reduce operating cost.Energy efficiency models are used to quantify this information and arrive at optimal solutions.Even at more tham 100 % capacity, the efficiency may be lower.
12/12/2011
Power Plant Performance Monitoring and Control
• This is a very powerful tool for Corporate / Unit level Energy Management for taking corporate and operational decisions at the right time and cost.
A typical MIS gives
• Energy efficiency of individual plant vs target
• Break-up of energy losses• Cost of power generation and control centres
• Transmission losses unitwise vs norms
• Specific power consumption etc
for corrective action at the Right time and Right Cost.
12/12/2011
Interfacing optimization model for steam pressure / temperature control.
Interfacing optimization model to superheated steam pressure and temperature control is feasible.In this case, the objective is to maintain constant steam pressure and a minimum superheated steam temperature (515 oC) to turbine inlet at the least time lag.The input parameter will be the flue gas temperature , flow rate of steam, inlet pressure/ temperature etc.
12/12/2011
Typical control profile
Control loop
Quench steam
12/12/2011
Questions on using Energy Efficiency Models
How effectively can these models be used in the absence of proper flow measurements for fuel and it’s properties ?
Indirect method is used for determining the efficiency of heaters / boilers etc which does not require flow data . Parameters used in the program are flue gas analysis , stack temperature , ambient temperature, Relative humidity and setting losses. This tallies very much with direct method of efficiency determination.
12/12/2011
Example - Equipment Maintenance / Replacement Decisions.
These models may be used for taking Equipment Maintenance / Replacement Decisions to minimise the Total Cost of Operation which determines the profitability and productivity of the industry.
020406080
100120140
I II III IV V
oprtg fixd total
12/12/2011
Failure Prediction of Equipment / components
These models in combination with Maintenance and Corrosion software may be used effectively to predict equipment / component failure due to corrosion, scaling, pitting, vibration etc using powerful models.
12/12/2011
Economics of Insulation
02468
1012141618
Rs '000000
1 2 3 4year
heat loss opr.cost savings
This approach evaluates Heat loss from various sections of the equipment, which could be used to identify the source of loss for taking insulation retrofit decisions.
12/12/2011
INNOVATIVE IDEAS
Cogeneration
Combined heat power cycle optimization
Trigeneration
Organic rankine cycle
Waste heat recovery
Using Energy efficient equipments etc