the genesys northwest model business/technologyinnovation...genesys builds a resource stack of...

The Genesys Northwest Model (Generation Evaluation System)

Bonneville Power Administration Hydro Modeling Conference

February 21-22, 2012


Outline

1. The Pacific Northwest Power Supply 2. What is GENESYS? 3. Uses for GENESYS 4. Additional Slides

BPA Hydro Conference

Pacific NW Power Supply

February 21-22, 2012 BPA Hydro Conference

Reservoir Storage & Runoff Volumes

0

50

100

150

200

Annual Runoff Storage

Volu

me

(Maf

)

USA Canada

Dry (78)

Avg (134)

Wet (193)


Variability in Hydroelectric Generation

0

5,000

10,000

15,000

20,000

25,000

Dry Average Wet Load

Aver

age

Meg

awat

ts

75% of NW electricity

on average

D E M A N D

D R Y

A VERA GE

W E T

51% of NW electricity driest year



GENESYS Northwest Monte Carlo Simulation of the NW Power Supply

Project-level Monthly Hydro Simulation

Hourly Economic Dispatch (including hydro)

Inter-regional Transmission Capacity (but not forced outages)

Random Variables: • Water Conditions

• Temperature/Loads

• Thermal Resource Forced Outage

• Wind



Transmission Modeled in GENESYS

NW region includes: East (E) West (W) Captain Jack (CJ)

Solid lines indicate transmission into and out of the region

C

E W

CJ

SW NC

SC



Genesys Overview For each game

For each year

Select random variables

For each hour

For each month

Water year

Temperature Year

Forced Outages

Build resource & load stack Do hourly dispatch

Build resource & load stack Do monthly dispatch


Wind Year

For each day

Get hydro peaking constraints Shape hydro generation

Monthly Dispatch

GENESYS builds a resource stack of available resources

Cheapest resources on the bottom Hydro is broken into blocks with

different prices GENESYS will dispatch enough

resources, starting with the cheapest, until all load is met



Hydro Blocks – Based on Value of Water (Not to Scale)

Very Cheap

Moderate

Expensive

Very Expensive

Emergency Only

Flood Control Curve

Assured Refill Curve

Actual Energy Regulation

Critical Rule Curve

Empty


Daily/ Hourly Dispatch

Daily – shape hydro energy to meet loads over 24 hours Ensure no peaking violations occur (see Trapezoidal model)

Hourly – build load and resource stacks Hourly hydro already assessed Dispatch cheapest resources first February 21-22, 2012 BPA Hydro Conference

Trapezoidal Model


NS

NP

NS

S0 S1 S2

NP = Number of Peak Hours NS = Number of Shoulder Hours S0 = Storage at beginning of off-peak period S1 = Storage at beginning of ramp up S2 = Storage at end of ramp down

Hours in the day

Meg

awat

ts

Objective Function & Constraints

Maximize on-peak generation Constraints Min and max total flow limits Max turbine flow limit Min required spill level End of week storage = beginning storage Lag time between projects


Peak vs. Energy Curve



Peak vs. Energy Curves for 1, 2, 4, 6, 8, 10 and 12 Hours

16000

18000

20000

22000

24000

26000

28000

10000 12000 14000 16000 18000Monthly Energy (MWa)

Cap

acity

(MW

)

1-Hr 2-Hr 4-Hr 6-Hr 8-Hr 10-Hr 12-Hr



January 1930

-5000

0

5000

10000

15000

20000

25000

30000

35000

40000

45000

1 27 53 79 105

131

157

183

209

235

261

287

313

339

365

391

417

443

469

495

521

547

573

599

625

651

677

703

729

Hour in Month

Meg

awat

ts

Net Demand NW Thermal NW Hydro Unserved Net Imports


Sample Simulated Dispatch

Uses for GENESYS

1. Adequacy assessments 2. Impacts of increasing amounts of wind 3. Resource cost effectiveness 4. Impacts of alternative hydro operations 5. Sub-year hydro generation forecasts 6. Climate change studies


Additional Slides


Effects of Wind on the Power Supply 1. Hydro peaking capability reduction due to

within-hour reserve requirements 2. Changes in thermal dispatch, more cycling and

more small scale variation in hydro generation 3. Impacts to spring oversupply conditions 4. Calculate the effective load carrying capability

of wind


Effects of Within-hour Wind Reserves on Hydro Capability INC Reserve – Generation dispatched during peak load when wind doesn’t blow DEC Reserve – Generation turned off during light load when wind does blow


Effects of INC and DEC Reserves on Hydroelectric Capability

8000

16000

24000

32000

0 5 10 15

Meg

awat

ts

Hour of Day

Orig w/DEC w/INC + DEC

DEC and INC both tend to flatten

hydro generating capability


Impacts to Resource Dispatch


0

5000

10000

15000

20000

25000

1 5 9 13

17

21

25

29

33

37

41

45

49

53

57

61

65

69

Meg

awat

ts

Hours in the Month

Simulated Dispatch – No Wind

Demand Therm - No Wind Hydro - No Wind

For Illustration Only


0

5000

10000

15000

20000

25000

1 5 9 13

17

21

25

29

33

37

41

45

49

53

57

61

65

69

Meg

awat

ts

Hours in the Month

Simulated Dispatch – 3K Wind Demand Therm - 3K Wind Hydro - 3K Wind Wind - 3K

Thermal absorbs the “energy” component while hydro absorbs the hourly variation



0

5000

10000

15000

20000

25000

1 5 9 13

17

21

25

29

33

37

41

45

49

53

57

61

65

69

Meg

awat

ts

Hours in the Month

Simulated Dispatch – 6K Wind

Demand Therm - 6K Wind Hydro - 6K Wind Wind - 6K

Thermal is OFF

Hydro hits min



Oversupply Conditions

Oversupply conditions occur when the minimum system generation exceeds firm load and secondary sales markets.


Feb Mar

Apr

May Jun

Jul

0

500

1000

1500

2000

2500

3000

0 2 4 6 8 10

Ove

rsup

ply

(MW

a)

Installed Wind (GW)

Oversupply in Average Megawatts (averaged over all hours of the month)

Oversupply conditions occurs even with no wind


No sales market assumed in this case




For an assumed intertie capacity N to S of 5,000 MW

Oversupply in this range

Oversupply Probability Curve

Effective Load Carrying Capability of Wind


What is ELCC? “Effective load carrying capability” is defined

as the amount of incremental load a resource can serve without degrading adequacy.

It is usually expressed as a percentage of a resource’s nameplate capacity.

ELCC is a function of the system the resource is added to – this is particularly important for wind.


Study Methodology Base case Remove all wind Calculate total annual average curtailment

Study cases Add 200 MWa of annual shaped load Add increments of wind capacity until the total

annual average curtailment equals that in the base

Wind data based on historic 2008-10 BPA wind fleet production

Repeat above with greater amounts of load


ELCC Results (+200 MWa load)

1200000

1250000

1300000

1350000

1400000

0% 10% 20% 30% 40% 50%

Tota

l Avg

Cur

tailm

ent (

MW

-hrs

)

Average Load/Wind Capacity (%)

Base Case

200 MWa Load, 500 MW Wind

200 MWa Load, 1000 MW Wind

ELCC = 26.4%



Annual Wind ELCC Results

0%

5%

10%

15%

20%

25%

30%

0 2000 4000 6000 8000 10000

Win

d EL

CC

(%)

Installed Wind Capacity (MW)

Average Incremental



Observations ELCC declines with increasing amounts of

wind because system flexibility is used up Eventually wind ELCC will flatten out Average annual wind generation is ~ 30%,

yet currently aggregate ELCC is ~ 18% Thus, can’t plan on average wind generation

Adding storage will increase ELCC Adding more diverse wind generation will

also increase aggregate ELCC


the genesys northwest model business/technologyinnovation...genesys builds a resource stack of...

Documents