Pacific NorthwestPacific NorthwestPacific NorthwestPacific NorthwestResource Adequacy StandardResource Adequacy Standardq yq y

John Fazio, NW Power and Conservation Council

IEEE RRPA Subcommittee

San Diego, CaliforniaSan Diego, California

July 27, 2012

OutlineOutline

• What is Resource Adequacy?q y

• The PNW Adequacy Standard• The PNW Adequacy Standard

• State of the System Report

• Current Assessment for the PNW

What is Resource Adequacy?What is Resource Adequacy?

Adequacy vs. ReliabilityAdequacy vs. Reliability• The North American Electric Reliability Corporation 

defines power system reliability to be composed of two basic and functional aspects of the electric system:

•• AdequacyAdequacy ‐ The ability of the electric system to supply the aggregate electrical demand and energythe aggregate electrical demand and energy requirements of the customers at all times, taking into account scheduled and reasonably expected 

h d l d f lunscheduled outages of system elements.

•• SecuritySecurity ‐ The ability of the electric system to withstand sudden disturbances such as electric short circuits orsudden disturbances such as electric short circuits or unanticipated loss of system elements.

•• Thus, adequacy is a part of reliabilityThus, adequacy is a part of reliability

What is an Adequacy Standard?What is an Adequacy Standard?q yq y

•• AdequacyAdequacy = having sufficient supplyq yq y g pp y

•• MetricMetric = a quantitative measure(of adequacy)(of adequacy)

•• ThresholdThreshold = a minimum level for an                 adequacy metricadequacy metric

•• StandardStandard = setting a threshold for an dadequacy metric

Forum’s Adequacy MetricsForum’s Adequacy Metrics

Metric Description

LOLPLoss of load probability = number of games with a problem divided by the total number of gamesLOLP a problem divided by the total number of games

Use of StandbyNumber of games that dispatch standby resources at least once divided by total gamesConditional value at risk = average annual

CVaR (energy) Conditional value at risk = average annual curtailment for 5% worst games

CV R ( k)Conditional value at risk = average single-hour

il f % fCVaR (peak) curtailment for worst 5% of games

EUEExpected unserved energy = total curtailmentdivided by the total number of games

LOLHLoss of load hours= total number of hours of curtailment divided by total number of gamesLoss of load expectation = total number of

6

LOLEp

events divided by total number of games

Sample Adequacy ThresholdsSample Adequacy Thresholds

Metric Threshold

LOLP 5 percent (used in the PNW)

Use of Standby Not commonly used

CVaR (energy) No common threshold

CVaR (peak) No common thresholdCVaR (peak) No common threshold

EUE No common threshold

LOLH R t 2 4 h /LOLH Ranges up to 2.4 hours/year

LOLE 1 event/10 years or 0.1 event/year

7

Physical vs. Economic AdequacyPhysical vs. Economic Adequacy

•• PhysicalPhysical = “keeping the lights on”

•• EconomicEconomic = “keeping rates low and minimizing annual price fluctuations”

• Economic adequacy is generally a higher standard (e.g. more controllable supply)

• Economic adequacy is usually targeted in integrated resource plans (IRP)resource plans (IRP)

• Physical adequacy is an “early warning” when resource development falls short for unexpected reasons

The PNW Adequacy StandardThe PNW Adequacy Standard

The Pacific NW Adequacy StandardThe Pacific NW Adequacy Standardq yq y

• Is a physical standardp y

• Assumes no transmission outages

• Uses the LOLP metric• Uses the LOLP metric

• Sets a 5% threshold for the LOLP

• Assess adequacy 5 years out,assuming existing or expected resources (including conservation)

Interpretation of the StandardInterpretation of the Standardpp• The likelihood of having at least one curtailment** five 

years into the future must be 5% or less for the power y psupply to be deemed adequate.

• Intended to be an early warning should resource development fall• Intended to be an early warning should resource development fall dangerously short

• Does not take economic factors into consideration• Not intended to be a resource needs assessment but could be usedNot intended to be a resource needs assessment but could be used 

to support one• Used as a safeguard in the Council’s resource strategy model  

** This represents a simulated curtailment. In reality, it represents the likelihood of having to take extraordinary measures to continue to provide service.

Adequacy Standard Methodology (1)Adequacy Standard Methodology (1)Adequacy Standard Methodology (1)Adequacy Standard Methodology (1)

• Use a chronological hourly Monte Carlo simulation computer model (GENESYS)

• Run many simulations (games) with different values for future unknown variables

• Future unknown variables include:• Water supply• Temperature (load) variation• Wind generation• Wind generation• Forced outages

Adequacy Standard Methodology (2)Adequacy Standard Methodology (2)Adequacy Standard Methodology (2)Adequacy Standard Methodology (2)• Model the transmission capacities of:p

• East‐west regional interties• NW to SW interties• NW to Canada interties

• Include an amount of market supply that pp ywe are reasonably sure will be available 

• Simulate the operation over every hour of every month

Adequacy Standard Methodology (3)Adequacy Standard Methodology (3)Adequacy Standard Methodology (3) Adequacy Standard Methodology (3) • Any game in which at least one curtailment 

i id d “b d”occurs is considered a “bad” game• LOLP = number of bad games divided by the 

total number of gamestotal number of games• The number of curtailments per game and the 

magnitude of curtailments do notdo not affect LOLPmagnitude of curtailments do not do not affect LOLP• A State of the System report provides more 

detailed information about the power supplydetailed information about the power supply, which includes frequency and magnitude  

The State of the System ReportThe State of the System Reporty py p

The State of the System ReportThe State of the System ReportThe State of the System ReportThe State of the System Report

• LOLP value and adequacy statusq y

• Values for additional adequacy metrics

• Monthly breakdown of LOLP• Monthly breakdown of LOLP

• Monthly and hourly use of resources and market supply

• Curtailment statistics 

DRAFT DRAFT Adequacy Assessment for 2015Adequacy Assessment for 2015(Not for Distribution)(Not for Distribution)(Not for Distribution)(Not for Distribution)

Metric Value Units Threshold

LOLP 0.93% Percent 5%

Use of Standby 1 1% Percent N/AUse of Standby 1.1% Percent N/A

CVaR (energy) 6,968 MW-hours N/A

CVaR (peak) 516 MW N/A

EUE 348 MW-hours N/A

LOLH 0.3 Hours/year 0.8 to 2.4

LOLE 0.02 Events/year 0.1

17

Total Annual Total Annual EnergyEnergy Curtailment Curtailment b bili *b bili *Probability Curve*Probability Curve*

150000

200000

Resulting Energy LOLP (after dispatching standby resources) = 0 07%

100000

MW‐Hou

rs standby resources) = 0.07%

Standby Resources

0

50000

M

0.0% 0.5% 1.0% 1.5% 2.0%Probability of Exceeding

18

*Prior to dispatching standby resourcesStandby resource energy = 83,000 MW‐hours per year

Highest Highest SingleSingle‐‐Hour Curtailment Hour Curtailment b bili *b bili *Probability Curve*Probability Curve*

8000

10000Resulting Capacity LOLP (after dispatching 

db ) b %

4000

6000

MW

standby resources) = about 1%

0

2000

000

Standby Resources00.0% 0.5% 1.0% 1.5% 2.0%

Probability of Exceeding

19

*Prior to dispatching standby resourcesStandby resource capacity = 662 MW in winter and 722 MW in summer 

Monthly LOLPMonthly LOLP(M h d i d d l )(M h d i d d l )(Months are treated independently)(Months are treated independently)

0.8

1.0

0.6

cent

Likely False Positives

0 2

0.4Perc

0.0

0.2

20

Oct

Nov Dec Jan

Feb

Mar

Apr

May Jun Jul

Aug Sep

Ann

Market Purchases by MonthMarket Purchases by Monthyy3000

90th PercentileOct‐Apr Max = 6,600 MW/Hour

2000

250090th  Percentile

Avg

1500

2000

W‐M

onths

May‐Sep Max = 2,000 MW/Hour

1000

MW

y p , /(1,000 MW is during LLH)

0

500

21

Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep

Average Monthly Average Monthly Market Purchase ProbabilityMarket Purchase ProbabilityMarket Purchase ProbabilityMarket Purchase Probability

4500

5000OctOct‐Apr Max   = 6,600 MW/Hour

3500

4000

4500

s

Nov

Dec

Jan

p , /

May‐Sep Max = 2,000 MW/Hour(1,000 MW is during LLH)

2000

2500

3000

MW‐m

onths Jan

Feb

Mar

A

500

1000

1500

M Apr

May

Jun

0

500

0% 20% 40% 60% 80% 100%b b l f d

Jul

Aug

Sep

22

Probability of Exceeding

Market Purchases by HourMarket Purchases by HourMarket Purchases by HourMarket Purchases by Hour7000

Oct

5000

6000

s

Oct

Nov

Dec

Oct‐Apr Max    = 6,600 MW

May‐Sep Max  = 2,000 MW( d )

3000

4000

MW‐Hou

rs Jan

Feb

Mar

(1,000 MW is during LLH)

1000

2000

M Mar

Apr

May

00 20 40 60 80 100

P b bilit

Jun

Jul

Aug

23

Probabilityg

Curtailment StatisticsCurtailment StatisticsStatistic Value Unit

Expected Events/year 0.02 Events/year

Average Event Duration 15 HoursAverage Event Duration 15 Hours

Average Event Magnitude 21,206 MW‐hours

Average Event Peak 1,697 MW 

Expected Curt Hours/year 0.3 Hours/year

% of Games with Curtailment* 1 1% %% of Games with Curtailment* 1.1% %

24**Prior to dispatch of standby resources

Conditions during an August Curtailment EventConditions during an August Curtailment Event

16% FORduring event

Ann Avg30%Low Water

25

DraftDraft Adequacy Assessment for 2015Adequacy Assessment for 2015(N t f Di t ib ti )(N t f Di t ib ti )(Not for Distribution)(Not for Distribution)

•• Supply is adequateSupply is adequate (LOLP < 5%)Supply is adequate Supply is adequate (LOLP < 5%)• LOLP dominated by peak events (i.e. not energy) • Most critical months are Dec, Jan and Aug , g• Adequacy depends on fair amount of market purchases

• December average = 23% of total market • August average       = 30% of total market

• However, full amount of market purchases are made less than 1% of the timeless than 1% of the time 

• System becomes inadequate when the combined increase in peak load and decrease in efficiency is about 2,000 MW 

26