wind%diesel%201 · 2017-07-12 · generator.thisincreasedamountof generaonreversesthepowerﬂow...

Wind Diesel 201 Rich Stromberg/Josh Cra; Alaska Energy Authority

IGRC Wind Power Conference -‐ Mar 2015

Intended Audience §  People who stayed awake through Wind Diesel 101

§  UKlity managers, staff seeking greater technical knowledge

§  Design/engineering firms

§  Project developers

§  Component suppliers

§  Neophytes interested in learning more about wind energy

2

Wake Losses §  The space behind a wind turbine that is marked by decreased wind power capacity due to the fact that the turbine itself consumes some of the energy in turning the rotor. The wind behind the turbine, in its wake, is less effecKve at generaKng energy for a certain distance in the downwind direcKon due to turbulence created by the upwind machine. Thus, when siKng a wind farm, it is important to space turbines as to minimize the impact each has on the others’ power producKon capacity, taking into account addiKonal costs for laying of electrical cable and other infrastructure required when machines are spaced further apart. (hXp://www.windustry.org/resources/wake-‐losses)

3

Horns Rev offshore wind farm -‐ Denmark Horns Rev 1 owned by VaXenfall. Photographer ChrisKan Steiness

4

Wake effect

Sandia Labs/ SWiFT Facility

5

Graphical representaKon of wind turbine wakes §  hXp://www.nvidia.com

hXp://www.eps.ee.kth.se/windpower/images/wakesim.jpg

hXp://www.windpowerengineering.com/construcKon/simulaKon/seeing-‐the-‐unseeable-‐in-‐a-‐rotor-‐wake/

6

Wake and the Park Effect §  Ideally, we would space turbines as far apart as possible in the prevailing wind direcKon. But land use and the cost of connecKng wind turbines to the electrical grid would indicate spacing them closer together.

§  As a rule of thumb, turbines in wind parks are usually spaced somewhere between 5 and 9 rotor diameters apart in the prevailing wind direcKon, and between 3 and 5 diameters apart in the direcKon perpendicular to the prevailing winds.

§  Typical park losses for properly spaced turbines are ~ 5%.

7

Seasonal changes in wind resources § One year of quality wind data is the minimum required to assess the local wind resource.

§ MulKple years give a beXer representaKon of variaKon and the potenKal resource.

§  Secondary load systems can be beXer designed with mulKple years of data.

§ Move forward with a project design, but leave the met tower up to improve project confidence.

8

AWS Truepower Wind Speed Anomaly Map Q3 2013 vs. Q4 2014

9

Turbulence

10

§  Turbulence induces addiKonal mechanical and vibraKon loads on wind turbines.

§  IEC61400-‐1 ediKon 3 defines the representaKve turbulence intensity as the mean + 1.28 Kmes standard deviaKon of random ten-‐min measurements. (The calculaKon has changed from ediKon 2, so it is important to understand which formula is used.) Load cases are defined by the reference turbulence intensity Iref which is equal to the mean turbulence intensity at 15 m/s.

Turbulence •  The site turbulence intensity (TI) must be less than the turbine design TI in

the higher operating range of the wind turbine (from 60% of rated power wind speed up through the cut-‐out velocity).

11

IEC Wind Turbine Class §  It is criKcal to know what the expected maximum wind speeds are at your turbine site.

§  Some turbines are designed for surviving high winds while others are designed to capture the most energy in calmer regimes.

§  Ensure that your turbine can survive the environment while producing the most energy possible.

§  Max average wind speed values were dropped in Ed. 3. Now the limit is 20% of Vref.

12

IEC Wind Turbine Class §  The flow inclinaKon must not exceed ± 8 degree for any wind direcKon.

§  The site average verKcal wind shear at hub height must neither be negaKve nor too steep. IEC 61400-‐1 express these condiKons as 0<a<0.2, where a is the exponent of a power law approximaKon near hub height.

13

Wind shear and surface roughness §  In general, the more pronounced the roughness of the earth's surface, the more the wind will be slowed down.

§  In the wind industry, people usually refer to roughness classes or roughness lengths, when they evaluate wind condiKons in a landscape. A high roughness class of 3 to 4 refers to landscapes with many trees and buildings, while a sea surface is in roughness class 0.

§  Concrete runways in airports are in roughness class 0.5.

15

Wind shear and roughness §  Roughness and wind shear are directly correlated.

§  This graph shows how wind speeds vary in roughness class 2 (agricultural land with some houses and sheltering hedgerows with some 500 m intervals), if we assume that the wind is blowing at 10 m/s at a height of 100 meters.

16

Wind shear – Logarithmic formula §  The wind speed at a certain height above ground level is:

v = v ref ln(z/z0 )/ln(z ref /z0 )

v = wind speed at height z above ground level.

v ref = reference speed, i.e. a wind speed we already know at height z ref . ln(...) is the natural logarithm funcKon.

z = height above ground level for the desired velocity, v.

z0 = roughness length in the current wind direcKon. z ref = reference height, i.e. the height where we know the exact wind speed v ref .

17

Roughness Class and Length

Rough-‐ ness Class Roughness Length m Energy Index (per cent) Landscape Type

0 0.0002 100 Water surface

0.5 0.0024 73 Completely open terrain with a smooth surface, e.g.concrete runways in airports, mowed grass, etc.

1 0.03 52 Open agricultural area without fences and hedgerows and very scattered buildings. Only softly rounded hills

1.5 0.055 45 Agricultural land with some houses and 8 metre tall sheltering hedgerows with a distance of approx. 1250 metres

2 0.1 39 Agricultural land with some houses and 8 metre tall sheltering hedgerows with a distance of approx. 500 metres

2.5 0.2 31 Agricultural land with many houses, shrubs and plants, or 8 metre tall sheltering hedgerows with a distance of approx. 250 metres

3 0.4 24 Villages, small towns, agricultural land with many or tall sheltering hedgerows, forests and very rough and uneven terrain

3.5 0.8 18 Larger cities with tall buildings

4 1.6 13 Very large cities with tall buildings and skycrapers

For example, assume we know that the wind is blowing at 7.7 m/s at 20 m height. We wish to know the wind speed at 60 m height. If the roughness length is 0.1 m, then

v ref = 7.7

z = 60

z0 = 0.1

z ref = 20

Therefore:

v = 7.7 ln(60/0.1) / ln(20/0.1) =

9.2966 m/s

18

Wind shear – Power law §  The power law exponent is α. For fairly flat terrain, it is common to use the one-‐seventh power law, where α = 1/7.

§  1/7 power law for height adjustments

for a known wind speed V1 at height H1, you can calculate V2 at height H2:

V2=V1*(h2/h1)(1/7)

§  For example: 9.008=7.7*(60/20)(1/7)

Terrain DescripOon Power law exponent, α

Smooth, hard ground, lake or ocean 0.10

Short grass on unKlled ground 0.14

Level country with foot-‐high grass, occasional tree 0.16

Tall row crops, hedges, a few trees 0.20

Many trees and occasional buildings 0.22 – 0.24

Wooded country – small towns and suburbs 0.28 – 0.30

Urban areas with tall buildings 0.4

19

So, with wind shear you can predict the wind speed at higher elevaOons…sort of.

§  EsKmaKng performance for a turbine with a 40-‐meter hub height off of 20m and 30m anemometers has lower risk than esKmaKng the performance of a turbine with a hub height of 70 or 80 meters.

§ Wind shear formulas esKmate annual averages – not diurnal paXerns.

20

Delta Wind Farm Diurnal PaTern

21

West TX A&M Diurnal PaTern

22

Drivers of winds at different heights §  Lower level winds are driven by solar heaKng of the Earth’s surface, so winds increase throughout the day and subside at night.

§ Higher-‐level winds are dominated by stably straKfied flows that sink down at night into the rotor swept area, but get pushed higher during the day as solar-‐induced turbulence picks up.

§  Knowing the true diurnal paTern at your hub height is criOcal when designing secondary load systems on medium and high penetraOon wind-‐diesel systems.

23

30m data extrapolated to 75m

24

75m data

25


26

75m data

27


28

75m data

29


30

75m data Big change in secondary load consideraOons

31

Wind Shade §  The higher you are above the top of the upwind obstacle, the less wind shade. The wind shade, however, may extend to up to five 10 Kmes the height of the obstacle at a certain distance.

§  If the obstacle is taller than half the hub height, the results are more uncertain, because the detailed geometry of the obstacle, (e.g. differing slopes of the roof on buildings, different species of bushes/trees) will affect the result.

32

Emmonak Wind Turbine Site

33

Wind Shade Calculator hTp://www.moOva.fi/myllarin_tuulivoima/windpower%20web/en/tour/wres/shelter/index.htm

34


35

Cape Stebbins

36


37


§ 

38

Wind Shade Calculator – 37m Turbine hTp://www.moOva.fi/myllarin_tuulivoima/windpower%20web/en/tour/wres/shelter/index.htm

Don’t forget to consider rotor diameter.

39


40


41

Cape Stebbins Preferred Turbine Site

42


43

Icing – hoar frost § Hoar frost forms in calm or very low wind periods when a volume of air cools below the point where it can hold the moisture. The frost nucleates on surfaces such as tree branches, anemometers and wind turbines. The crystals are light, fluffy.

44

Hoar frost

45

Hoar frost

46

It only takes a little wind to knock off hoar frost and let anemometers resume spinning.

How do we treat hoar frost signals in our wind resource analysis?

§  Take no acKon. Leave the data points in place. They accurately represent what the wind is doing. Do not delete. Do not synthesize.

47

Stall-‐regulated turbine icing

Turbine output in kW(right axis on graph)

Both std met tower anemometers and heated turbine anemometers track well

Turbine output lags even after std anemometers have deiced

Stall-‐regulated turbine power curve

Icing – heavy rime ice

§  Rime is a rough white ice deposit which forms on surfaces exposed to the wind. It is formed by super-‐cooled water droplets of fog freezing on contact with a surface as it dri;s past. Rime ice can grow to become quite heavy.

50

Heavy rime ice – stall regulated wind turbine

51

Heavy rime ice prevents proper lift.

Heavy rime ice – pitch regulated wind turbine

52

How do we treat rime ice signals in our wind resource analysis?

§  Take no acKon. Leave the data points in place. They accurately represent what the wind turbine is likely to do and we can’t disKnguish between hoar frost and rime ice. Do not delete. Do not synthesize. Forecast error will be slight, but conservaKve.

53

Voltage Rise in Distributed GeneraOon (DG) Systems

“ConnecKons of distributed generaKon (DG) in distribuKon networks are increasing. These connecKons of distributed generaKon cause voltage rise in the distribuKon network.” -‐ hXp://seit.unsw.adfa.edu.au/staff/sites/hrp/papers/mhp11a-‐c.pdf Analysis of Voltage Rise Effect on DistribuKon Network with Distributed GeneraKon M. A. Mahmud, M. J. Hossain, H. R. Pota

“Since the modern distribuKon systems are designed to accept bulk power from the transmission network and to distribute it to customers, the flow of both real and reacKve power is always from the higher to lower voltage levels. However, with significant penetraKon of distributed generaKon, the power flows may become reversed and the distribuKon network is no longer a passive circuit supplying loads but an acKve system with power flows and voltages determined by the generaKon as well as load.”

“ConnecKons of distributed generaKon in distribuKon systems are suscepKble to voltage rise. Moreover, the impact of losing a single or a few distributed generaKon following a remote fault may not be significant issue, but the connecKon or disconnecKon of a large penetraKon of distributed generaKon may become problemaKc which may lead to sudden appearance of hidden loads and affect the voltage profile of low voltage distribuKon network.”

54

DG Voltage Rise Example Analysis of Voltage Rise Effect on DistribuKon Network with Distributed GeneraKon M. A. Mahmud, M. J. Hossain, H. R. Pota

In this case, 240 KW generator 15km away from the primary distribuKon system is replaced by a 1 MW generator. This increased amount of generaKon reverses the power flow through the line, from the generator towards the DistribuKon System (DS).

The voltage profile of DS with 1 MW of distributed generaKon is shown in Fig. 7. From Fig. 7, it is seen that the voltage in some parts of the system rises above the permiXed +6% voltage limit. Note that <3% is preferred.

55

DG Voltage Rise Factors Analysis of Voltage Rise Effect on DistribuOon Network with Distributed GeneraOon M. A. Mahmud, M. J. Hossain, H. R. Pota

The level of DG generaKon that can be connected to the distribuKon system depends on the following factors:

§  voltage at the primary DS

§  voltage level of the receiving end

§  size of the conductors as well distance from the primary DS

§  load demand on the system

§  other generaKon on the system

56

DG Voltage Rise MiOgaOon Analysis of Voltage Rise Effect on DistribuOon Network with Distributed GeneraOon M. A. Mahmud, M. J. Hossain, H. R. Pota

The voltage rise on DS can be miKgated through the following approaches:

§  Resistance reducKon (increase conductor size or energize to higher voltage)

§  ReacKve power compensaKon (switched capacitor or DVAR)

§  Coordinated voltage control

§ GeneraKon curtailment

57

DG Voltage Rise – Other Reading “A Case Study of a Voltage Rise Problem Due to a Large Amount of Distributed GeneraKon on a Weak DistribuKon Network” – Sami Repo, et al. hXp://labplan.ufsc.br/congressos/PowerTech/papers/51.pdf

“The integraKon of relaKvely large capacity of wind power into a weak distribuKon network may cause a voltage rise problem during low demand periods.”

“A review on voltage control methods for acKve distribuKon networks” – Tengku Hashim, et al hXp://pe.org.pl/arKcles/2012/6/71.pdf

“The convenKonal distribuKon networks are designed based on the assumpKon of unidirecKonal power flow. With the increasing connecKon of DG, the network has become more dynamic with bidirecKonal power flow and it[sic] known as acKve distribuKon networks (ADN).”

“With the increasing number of DG penetraKon, the issue of voltage level in distribuKon systems has become important. Increasing the number of connected generators will result in voltage rise above its permissible level.”

58

DG Voltage Rise – Other Reading “IntegraKon of Distributed GeneraKon in Low Voltage Networks: Power Quality and Economics” – KonstanKnos Angelopoulos hXp://www.esru.strath.ac.uk/Documents/MSc_2004/angelopoulos.pdf

“It is possible to esKmate the effect of a generator by using the standard voltage drop equaKons with reverse power flow. The voltage drop along a feeder due to a load is approximately equal to:

Vdrop = IRR+IXX

Where:

Vdrop= voltage drop along the feeder

R = line resistance, ohms

X = line reactance, ohms

IR = line current due to real power flow, amps (negaKve for a generator injecKng power)

IX = line current due to reacKve power flow, amps (negaKve for a capacitor)

59

DG Voltage Rise Analysis on Alaska WD Systems

§  Power flow analysis can be costly and take Kme, but is needed in come cases.

§ UVIG DG toolkit is a quick method to determine if more detailed PF analysis is needed. hXp://www.uwig.org/distwind/default.htm

§ A simple voltage drop/rise calculaKon can be done in two minutes.

60

Voltage Rise – Kotzebue Example

61

Single phase VD = (2 * L * R * I) / 1000 ftDistance in miles 4Equivalent feet 21,120Resistance in Ohms/1,000 feet from chart at right 0.1265 2/0 QuailLoad in amps is based on total power and line voltageMax power (Watts) from all wind turbines 1,100,000

Voltage rating of transmission line 12470Single phase amps from wind turbine 88.21Convert to 3-‐phase (Div by sqrt of 3) gives load in amps from turbine 50.93Using above bold formula, voltage drop/rise is -‐-‐-‐-‐-‐-‐> 272.14Percentage of voltage drop/rise 2.18%

3-phase VD = SPVD * (1.732/2) Drop between any 2 phases3-‐phase voltage drop/rise is-‐-‐-‐-‐-‐-‐-‐-‐-‐-‐-‐-‐-‐-‐-‐-‐-‐-‐-‐-‐-‐-‐-‐-‐-‐-‐-‐-‐-‐-‐> 235.68Percentage of voltage drop/rise 1.89%

Voltage Rise – Kotzebue Example

<3% is desired

Before adding two EWT 900kW turbines.

62

Voltage Rise – Kotzebue Example Single phase VD = (2 * L * R * I) / 1000 ftDistance in miles 4Equivalent feet 21,120Resistance in Ohms/1,000 feet from chart at right 0.1265 2/0 QuailLoad in amps is based on total power and line voltageMax power (Watts) from all wind turbines 2,900,000



0.622596<3% is desired

Voltage can rise as wind power increases on distributed generation microgrids.

A;er adding two EWT 900kW turbines.

63

Voltage Rise at KEA A;er 2013 Energize to 25kV

Goal achieved

Single phase VD = (2 * L * R * I) / 1000 ftDistance in miles 4Equivalent feet 21,120Resistance in Ohms/1,000 feet from chart at right 0.1265 2/0 QuailLoad in amps is based on total power and line voltageMax power (Watts) from all wind turbines 2,900,000



0.622596

64

InducOon Generators vs. Inverter Systems

§  An inducKon (asynchronous) generator must have its magneKc field maintained through the same mechanism as an inducKon motor. It must exchange energy with a capacitor or with a synchronous generator that can be adjusted to “act as a capacitor.” In order to funcKon as a generator, an inducKon generator requires an external source of reacKve volt-‐amperes (VARs). This is typically supplied by the diesel gensets. Power factor drops as the WTG produces more energy.

§  Inverter-‐based WTG controllers create a wall from the microgrid where VARs are produced by the inverter using power from the wind turbine once it has spun-‐up. The microgrid only sees clean power.

65

Generator Sizing and Spinning Reserve §  Being able to step up or down to the appropriate size diesel genset as wind producKon moves up and down can minimize fuel efficiency hit.

§  Larger diesel genset may sKll be needed for VARs support or spinning reserve.

§  Sufficient spinning reserve (diesel, baXery, etc.) must be maintained to handle sudden drops in wind output. 50% of current wind energy may be needed.

§ Diesel generators will see a greater number of starts/stops when wind energy is introduced – some efficiency loss.

66

Generator Sizing and Spinning Reserve Engine Make/Model

Serial #Min Load %

Rated Capacity (kW) (kVA)

Average Load on Genset

Average Load on Genset w/ Wind

Det diesel 60 363

Cummins KTA 19G4 499

MTU 12V2000 700

Heat Recovery Loop: None currently, but possibility for water treatament plant and the school.

46% 38%

33% 27%

Application/Grant #

Diesel GensetsNoorvik Wind Farm

Comments: Manual switchgear in Noorvik would need to be upgraded and possibly new feeders.

64% 52%30%

30%

30%

67

What if the wind doesn’t drop off suddenly, but keeps gejng stronger?

68


69


§  If all turbines are set to trip off at exactly 25 m/s, Unalakleet could lose 600kW of power generaKon in a few seconds.

§  Which diesels are online and how quickly can they make up for the 600kW?

§  Staggering wind turbine cut-‐out speeds can minimize the power loss steps to 100 or 200kW.

§  Single wind turbines make this harder to accomplish unless they have variable pitch blades plus controls that allow for reducing energy output as the turbine gets close to the cutout speed.

§  Smart systems control logic will bring addiKonal spinning reserve online when wind turbines get close to cut-‐out speed.

70

12-‐month Unalakleet met tower study showed no incidents of hijng cut-‐out speed

However, UVEC has seen instances where wind turbines cut out at 25 m/s and the diesel gensets trip offline.

71

Another problem with strong winds §  A small community with small load in a class 7 wind regime.

§  Average load is 29kW. Average wind penetraKon is 81%.

§  One 65kW wind turbine installed – stall-‐regulated, basic controller.

§  Turning on the wind turbine at 15-‐25m/s first causes an in rush of current into the wind turbine’s inducKon generator. Then, the turbine pushes 65kW of power back onto the local grid.

§  If diesel genset and secondary loads can’t respond fast enough, high voltage or frequency will trip off the diesel genset and village loses power.

SoluKons:

A single smaller turbine.

MulKple smaller turbines with automated switchgear that turns on one turbine at a Kme.

Develop a smart wind turbine controller that starts the turbine with a long ramp rate to max power.

72

Other wind turbine features to consider in your system

§  So; start

§ Dynamic braking

§ Variable-‐pitch blades

§  Tilt-‐up towers vs. monopole towers vs. la}ce towers

73

Secondary load consideraKons

74


75

Secondary load consideraOons §  Is there a heat recovery loop on the exisKng diesel system?

§ How much energy (mmBTUs) currently goes into the HR loop and at what rate throughout the year?

§ How much energy is pulled off the HR loop by value loads and non-‐value loads? At what rate throughout the year?

76

Secondary load consideraOons §  Does the “dead zone” where wind picks up and diesels throXle back reduce the energy in the HR loop below the value load demand? If this happens fairly o;en, consider placing an electric boiler on the HR loop before any other secondary load opKons.

§  If the energy loss in the HR loop rarely or never drops below the value load demand, an electric boiler on the HR loop buys you no economic benefit for your excess electricity. You should consider value electric heat loads elsewhere in the community (school, village office, water treatment, washeteria, wastewater system, residenKal).

§  Don’t overlook the opportunity for dispatchable electric loads like pumping water.

77

Secondary load consideraOons § At what rate do your thermal loads “consume” heat (mmBTUs)?

§ Will your wind turbines produce excess energy at a rate faster than can be absorbed by your secondary thermal loads?

§  If so, you will either need to curtail wind turbines and lose economic benefit, send excess power to an open air dump load (no value) or add electric boilers/heaters to value loads elsewhere in your community.

78

§  Even though the health clinic uses almost 2X the amount of annual excess energy produced by the wind turbines, it is not large enough to handle all the excess energy in the system.

79


Community building/load Connected to HR Loop?

Current annual heating oil consumption*

Thermal mass -‐ Equiv. gals. of storage

MMBTU Equiv

kWh Equiv

Average kW

Design Day Heat

Public Works-‐HEMF Y 19,216 2,652 743,163 84.84 Suspect boiler setpoint set above level to gain benefit from HR loop. Estimate 20% of total is unmet.Sewer Plant Y 13,695 1,890 529,639 60.46 Estimate 20% of total load is unmetSchool N 116,800 16,118 4,517,240 515.67 1840000PSO N 6,348 876 245,502 28.03 100000 <BTU/HrHealth clinic N 14,219 1,962 549,925 62.78 224000 <BTU/HrWater plant N 11,426 1,577 441,904 50.45 180000 <BTU/HrFire Station N 16,758 2,313 648,126 73.99 264000 <BTU/HrPower plant Y 1,625 224 62,847 7.17 Estimate 20% of total load is unmet

0 0 0.000 0 0.00

Totals 200,087 27,612 7,738,346 883.37 331,107 <<Excess kWh from HOMER

OperaOons and Maintenance 1.  Power generaKon and distribuKon systems that are not

properly maintained eventually stop generaKng and distribuKng power.

2.  A poorly maintained petroleum-‐based energy system makes a poor foundaKon for a wind energy system.

3.  Wind energy systems must be well maintained in order to save you money.

4.  UKliKes that focus on operaKons, maintenance and training have successful wind-‐hybrid energy systems.

80

O&M key components §  SCADA system

§  System/error logs

§  PM checklists (daily/weekly/monthly/quarterly/annual)

§  Spare parts (storage space)

§  TroubleshooKng methods

§  Decision trees

§  Staffing

§  Training

81

O&M strategy §  Focus on preventaKve maintenance

§  Address outliers

§  Reduce sources of variaKon across your systems

§  Have wriXen procedures

§  Train to your procedures

§ When methods improve, update procedures

§  Take Kme for thorough root-‐cause analysis and tesKng of soluKons

§  The beXer your data, the beXer your data-‐driven decisions

82

Use data to idenKfy areas of focus

83


84


85


86


87


88

Conclusions § Much of the needed design acKvity on Alaska wind energy systems deals with integraKng wind power with the exisKng power plant, distribuKon system and community heat loads.

§ Detailed understanding of how your wind turbines will interact with your exisKng or planned power generaKon and distribuKon is key to a successful project that will last decades.

§ A long term commitment to staffing, training and O&M will ensure you get the most benefit from your energy system.

89

CongratulaOons! You’ve survived Wind-‐Diesel 101 and 201!

More quesKons?

Rich Stromberg, Wind Program Manager Tel. (907) 771-‐3053

E-‐mail: [email protected]

Josh Cra;, Asst. Project Manager -‐ RPSU

Tel. (907) 771-‐3043 E-‐mail: jcra;@aidea.org

90