optimal integration of steam turbines in …optimal integration of steam turbines in industrial...

34
OPTIMAL INTEGRATION OF STEAM TURBINES IN INDUSTRIAL PROCESS PLANTS J D Kumana, MS ChE Kumana & Associates, Houston, Tx [email protected] (281) 437-5906 AIChE-STS, network meeting Houston, 2 Feb 2018

Upload: others

Post on 17-Mar-2020

32 views

Category:

Documents


1 download

TRANSCRIPT

OPTIMAL INTEGRATION

OF STEAM TURBINES IN

INDUSTRIAL PROCESS PLANTS

J D Kumana, MS ChE

Kumana & Associates, Houston, Tx

[email protected] (281) 437-5906

AIChE-STS, network meeting

Houston, 2 Feb 2018

Kumana & Associates © 2018 2

Outline

• Definitions: CHP and Efficiency

• Thermodynamics Review

• Energy integration theory

• CHP models

Kumana & Associates © 2018 3

• CHP = Combined Heat and Power (= energy utility system for the plant site)

• Steam Turbines are Heat Engines that operate on the Rankine cycle. They convert DP into Shaftwork; a generator then converts Shaftwork into Elec power

• Thermodynamic Efficiency is defined as

• For Generation, 1 useful output = Power only. Machine eff= ~20%, System Eff = ~35%

• For Cogeneration, 2 useful outputs = Power + Process Heat, Machine eff = ~20%, but System Eff ~75-80%

Useful Energy Output

Energy Input

DEFINITIONS: CHP & EFFICIENCY

Kumana & Associates © 2018 4

This is CHP, but not Cogeneration

BOILERS

CONDENSER

PROCESS

FUEL

KWSTEAM

TURBINE

BOILER EFF ~ 80%

POWER GEN EFF < 25%

LATENT HEAT OF ST EXHAUST IS WASTED

STEAM

Pure

power

gen

Kumana & Associates © 2018 5

This is both CHP and “Co-Generation”

BOILERS

PROCESS

FUEL

KWSTEAM

TURBINE

LP STEAM

OVERALL EFF ~ 75%

LAT HT OF EXHAUST STM IS USED IN THE PROCESS

Kumana & Associates © 2018 6

Alternative Cogen configurations

Extraction Turbine Induction Turbine

Kumana & Associates © 2018 7

Variations – hybrid Cogen and Condensing

Extraction turbine Induction turbine

Kumana & Associates © 2018 8

Simple Rankine Cycle flowsheet

Schematic shown is for cogeneration mode

Kumana & Associates © 2018 9

Difficult to match Heat:Power ratio of process

4 Basic Configs – which do you think is most efficient?

Most

efficient

Kumana & Associates © 2018 10

The ultimate Combined-cycle Cogen scheme

PROCESS

AIR GAS

EXHAUST

TO ATMOS

GAS TURBINE

HRSG

KW

HP STEAM

KW

LP STEAM

OVERALL EFF ~ 85%

ELEC EXPORT

Kumana & Associates © 2018 11

Different types of ST Efficiency

• Machine Efficiency = W/Qin = (H1-H2)/H1

• Isentropic Efficiency = W/[M.(H1–H2)max] = (H1-H2)/(H1-H’2)

• System efficiency

M, P2, T2, H2

W

PROCESS

Equipt

M, P1, T1, H1

M - mm

1

2

M.H

Hmm).-(M 3413.kW 2.

H’2 = exhaust vapor enthalpy IF the

expansion were isentropic (which it

is not, and can never be)

Kumana & Associates © 2018 12

A Bit of History …

US Power plants stopped cogenerating ~1960

Kumana & Associates © 2018 13

THERMODYNAMICS REVIEW

Rankine cycle on the P-V diagram

P-V Diagram for Water

0

500

1000

1500

2000

2500

3000

3500

0.01 0.1 1 10

Specific Volume, ft3/lb

Pre

ssu

re, p

sia

1

3

4

2

Rankine Cycle

1. BFW pumpimg

2. Stm gen in Boiler

3. ST expansion

+ pow er gen

4. Condensation

Kumana & Associates © 2018 14

Power generation step (#3) on Mollier Chart

• Adiabatic expansion (from 600 psig, 700oF to 50 psig)

• Isentropic efficiency

Mollier Chart (H-S) for Steam

900

950

1000

1050

1100

1150

1200

1250

1300

1350

1400

1450

1500

1550

1600

1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2.0 2.1

Entropy, Btu/lb-F

En

tha

lpy, B

tu/lb

Saturation

Pr, psig

Quality, % stm

Temp, F

1200 600

300150

50

1.0

800

700

600

215

300

400

500

92% Quality

97 %

Saturation

line

--12 psig

1

2

Kumana & Associates © 2018 15

Effect of P2/P1 on Machine Efficiency (W/Qin)

Near-optimal

Inlet Conditions

for industrial

cogen systems

Theoretical Machine Efficiency tops out at ~13% for BPST and 24% for

CST before moisture content in turbine reaches dangerous levels.

Power-to-Heat Ratio vs Steam Pressure Ratio

0.00

0.05

0.10

0.15

0.20

0.25

0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8

P2/P1

Pow

er:

Heat

ratio

BPST

limit

Condensing

Turbine limit

Kumana & Associates © 2018 16

Effect of P2/P1 on System Efficiency

System Efficiency peaks when exhaust steam is saturated,

drops rapidly as P2/P1 is falls, slowly as P2/P1 rises

System Efficiency vs P2/P1 ratio

0%

20%

40%

60%

80%

100%

0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8

P2/P1

Syste

m e

ne

rgy E

ff

Exhaust

stm is dryExhaust

stm is wet

Condensation starts

at P2 = 53 psig

Kumana & Associates © 2018 17

Next: What is the Optimum Exhaust Pressure?

• P2 should be at a high enough pressure that it can be used for process heating

• If there are multiple steam levels in the process, an extraction type turbine should be considered, with both exhaust pressures above ambient.

• The amounts should match the process steam requirements ( “thermal match”)

• For higher P2 or W/Qin increase P1 and T1

PINCH ANALYIS provides the ANSWER

Kumana & Associates © 2018 18

OPTIMUM TURBINE INTEGRATION

Qcold

Qhot

Te

mp

Heat Load

Qhot & Qcold are the

energy targets

Pinch = minimum DT

reqd for ht tr “Pinch

Analysis”

It is possible to consolidate ALL the heating and cooling duties in the process

into two Composite Curves that show the enthalpy change requirements

between the entire temperature range over which the process operates

Kumana & Associates © 2018 19

The Pinch Principle - 1

If we allow XP heat transfer, Qh and Qc both increase by XP

QCmin+ XP

Source

QHmin+ XP

Sink

H

T

XP

Not economic

because DT< DTmin

?

Kumana & Associates © 2018 20

The Pinch Principle - 2

To achieve the Energy

Targets, DO NOT

• use Steam below Pinch

• use CW above Pinch

• transfer heat from

process streams above

Pinch to process

streams below Pinch

T

Cooling

Water

Steam

Process Heat

TransferPinch

Temp

Kumana & Associates © 2018 21

Steam Turbine Integration options

100% conversion of Q W

T

Heat

EngineW

Q - W

QA - (Q - W)

A + W

Heat

EngineW

Q

B - QQ - W

No improvement in system

T

Heat

EngineW

Q - W

B + (Q - W)

QA

A + Q

Kumana & Associates © 2018 22

Summary of Energy Balances

= Machine

efficiency

Kumana & Associates © 2018 23

Grand Composite Curve - GCC

T

H

REFRIGERATION

COOLING WATER

LP STEAM

HP STEAM

Used for

utilities

selection

Kumana & Associates © 2018 24

Correct Integration of Steam Turbine

• GCC shows us exactly how much HP and LP steam is needed, and the right P/T levels

• ST must alwaysexhaust ABOVE the Process Pinch

• When designed this way, payback is very good, typically 3-4 yrs

T

H

QLOSS

FuelW

HPS

LPS

Process

Pinch

Grand Composite

Curve

CW

Fuel = HPS + LPS + W + Qloss

Kumana & Associates © 2018 25

Total Site Source-Sink curves

Net process

cooling demand

= available heat

Net process

heating

demand

T

wHP

Req. Q

Enthalpy, MMBtu/h

CW

BFW

LP stm

Sink

Sink

Source

Kumana & Associates © 2018 26

Optimize Configuration

LP

MP

HPHP

MP

LP Sink

Fuel+

+

LP

MP

HPHP

MP

LP

Source

Fuel

+

LLP LLP

IP

+

+

Power generation

increased

Reduction in fuel

consumption

EXISTING

OPTIMIZED

Kumana & Associates © 2018 27

CHP SIMULATION MODELS

Kumana & Associates © 2018 28

Excellent Tool for Analysis

Model should include all Key System Features: Multiple steam levels

Multiple boilers (with eff. curves)

Process WHBs

Steam and Gas turbines (incl HRSG)

PRVs, Desuperheaters

Condensate recovery (by steam pr level)

Boiler blowdown flash & HX

Deaerators (could be > 1)

“Dump condenser”, if needed

Economizer for BFW preheat

BFW integration with process

Process power demand

Kumana & Associates © 2018 29

CHP Optimization Guidelines

• Set BPST exhaust pressures based onprocess steam headers (from GCC)

• Set steam flows through BPSTs based on process heating duties at each Pr level

• Condensing Turbines invariably a BAD idea

• Minimize flows through PRVs

• Use highest feasible DeAerator pressure(s)

• Maximize condensate recovery

• Preheat cold BFW makeup water by using it as a cooling medium in the process

Kumana & Associates © 2018 30

On-line Utilities Optimization

Hydrogen Fuel Steam Water ElectricityUtility Systems

External Utilities Contracts

Emissions Regulations

ProcessIndustrialSite

Real-Time Optimizer finds the best way to operate all utilities subject to contractual, environmental and operational constraints

OptimumUtilitiesOperationsReport

Measurements

OptimumSet Points

Key Performance IndicatorsMonitoringand AccountingReports

From VisualMesa® brochure, Courtesy of Soteica LLC, Houston, Tx

Kumana & Associates © 2018 31

Expected Benefits and Costs

• Typical savings = 3-5% of baseline (operator-

optimized) energy costs

• Typical installed cost = $500-900K

• Typical Payback << 1 yr

• Proven in dozens of Oil refineries, Chemical

plants, Pulp/Paper mills (can be deemed a

Best Practice)

Kumana & Associates © 2018 32

Case Study – Operate closer to OptimumOptimización Visual Mesa

Ahorros Anuales Predichos - Siguiendo las Sugerencias de Optimización

TODOS LOS DATOS RECOLECTADOS, HASTA EL PRESENTE

5,30 %

1,19%

-

1.00

2.00

3.00

4.00

5.00

6.00

7.00

8.00

9.00

10.00

19-12

26-12 2-1 9-1 16-1

23-1

30-1 6-2 13-2

20-2

27-2 6-3

Day

Saving

s / Tot

al Ener

gy Co

sts (%

)

Ahorros AnualesPromedio del Período Anterior a la OptimizaciónPromedio Después de Tomar Acciones de Optimización

> 2 MM €/year

> 4%

Y axis = Deviation from Optimum = Remaining Savings Opportunity

Graphic supplied by Soteica LLC, Houston, Tx

Kumana & Associates © 2018 33

IN CONCLUSION

• Use GCC to choose Stm Levels and Loads

• Use BPSTs in cogen mode when possible

• Condensing steam turbines are Invariably Bad*

• Use TSSS to identify optimum CHP structure

• Use CHP models to optimize parameters

• Always optimize process demand before trying to design/optimize the CHP system

• Ability to export excess power to the Grid at a fair price is critical to optimizing energy efficiency at National scale, and minimizing global GHG emissions

with a few rare exceptions

Kumana & Associates © 2018 34

Optimum Process Integration

It’s like a jig-

saw puzzle,

but well worth

the effort