Some issues on Scarcity Pricing* FERC Research in Market Oversight - - PowerPoint PPT Presentation

some issues on scarcity pricing
SMART_READER_LITE
LIVE PREVIEW

Some issues on Scarcity Pricing* FERC Research in Market Oversight - - PowerPoint PPT Presentation

Some issues on Scarcity Pricing* FERC Research in Market Oversight (RIMO) Program CPUC Project Bishu Chatterjee (CPUC) / Keith Collins (FERC) In collaboration with FERC Division of Energy Market Oversight August 10, 2007 CAISO Market


slide-1
SLIDE 1

1

Some issues on Scarcity Pricing*

FERC Research in Market Oversight (RIMO) Program CPUC Project Bishu Chatterjee (CPUC) / Keith Collins (FERC) In collaboration with FERC Division of Energy Market Oversight August 10, 2007 CAISO Market Surveillance Committee Meeting

* The views expressed in this project are solely those of the researchers and not necessarily those of the California Public Utilities Commission nor the Federal Energy Regulatory Commission

slide-2
SLIDE 2

2

Overview

  • FERC RIMO Program
  • Scarcity Pricing context in CAISO
  • Review of Scarcity Pricing in RTO/ISO
  • Examine scarcity conditions based on

predefined Heat Rates

  • Examine Off-peak hour shortage

conditions

  • Role of Scarcity Pricing in energy only

market versus capacity market

  • Conclusion
slide-3
SLIDE 3

3

The Scarcity Pricing context

In MRTU Release 1 the CAISO will have a limited scarcity pricing mechanism where bids will be raised to bid cap when there are insufficient energy bids In September 2006 MRTU Order, FERC ordered a more comprehensive reserve shortage scarcity pricing to be implemented within 12 months of the MRTU Release 1 In April 2007 MRTU Order FERC further emphasized that 1) Prices should rise when energy and reserves are short in Day Ahead and Real Time 2) Reserve shortage scarcity pricing that applies administratively-determined graduated prices to various levels of reserve shortage

slide-4
SLIDE 4

4

Scarcity Pricing (SP)

SP is based on the idea that under scarcity conditions generating units will receive higher compensation. This additional revenue stream helps to incentivize investment in new generation and promotes

  • verall system reliability. When supply

conditions are tight and drop below a pre- determined threshold level, the price for additional MW significantly rises.

slide-5
SLIDE 5

5

Scarcity Pricing (SP) Program and Scarcity Pricing

The SP exists in various ISO /RTOs can be grouped as follows: SP Program: SP Program has pre-specified prices/rules so when scarcity conditions (a tight supply situation due to pre-specified shortage) are triggered, such pre- administratively determined prices set SP. SP Programs exist in PJM, MISO, NYISO, ISONE, CAISO (MRTU Release 1A).

slide-6
SLIDE 6

6

SP Programs

  • Reserve demand Curve for shortage pricing uses

pre-defined shadow prices for reserves under scarcity conditions and energy prices automatically reflect cost of reserves under scarcity conditions. Example, NYISO and ISONE.

  • SP Program also determines SP via the co-
  • ptimization of energy and ancillary services. The

locational and temporal constraints for ancillary services given congestion and scarcity determine

  • SP. Under the co-optimization, energy and

reserves are priced simultaneously to reflect the

  • pportunity cost of not providing the product.

Example, NYISO and ISONE. PJM optimizes energy and reserves sequentially.

slide-7
SLIDE 7

7

SP via demand/supply conditions that are

  • utside the SP Program

SP can also be set outside the SP program by demand and supply conditions due to tight supply conditions. Such SP do not have pre- specified prices but are based on willingness of load to curtail load at such price. These concepts are similar to Value of Lost Load w/ an Offer Cap. Example: ERCOT ($1500/MWh), PJM, MISO, NYISO, ISONE ($1000/MWh),CAISO( $500/MWh).

slide-8
SLIDE 8

8

SP in PJM

SP Trigger ( SP region is incapable of meeting demand under normal economic dispatch conditions) and PJM performs any of the following:

  • 1. Distribution Level Voltage reduction action by 5

percent to provide load relief

  • 2. Dispatch of online generators partially

designated as Maximum Emergency into emergency output levels, entirely designated as Maximum Emergency above their minimum load points, and offline generators designated to run

  • nly in emergency conditions.
  • 3. Manual Load Dump – to disconnect firm

customer load

slide-9
SLIDE 9

9

SP in PJM…(2)

SP = Highest Offer Price to supply either energy or reserves in Real-Time subject to the overall offer cap of $1000/MWh In the State of the Market Report 2006 the PJM Market Monitor recommends

  • Stages of SP – based on system conditions with

progressive impacts on prices

  • Locational Price Signals
slide-10
SLIDE 10

10

SP in MISO

SP Trigger

  • 1. Day Ahead Maximum Generation Emergency

conditions (event steps 1 to 5) – shortage when the sum of demand bids (including price sensitive demand), exports, and virtual bids cannot be satisfied from generation, imports, and virtual

  • supply. [market demand ≥ market supply].
  • 2. Real Time shortage when real Time demand

forecast ≥ Available generation + Self Schedules + Demand Response Resources. SP = The higher of the Market Participant submitted Resource Offer/ARC in step 1 or $1000/MWh in step 2

slide-11
SLIDE 11

11

Step 1 SP in MISO…(2)

Step 1 uses up to 50% of the emergency range of online generators to handle load change less than 1 hour in duration (addresses temporary supply and demand) Adequate Ramp Capability (ARC): ARC captures the value

  • f operational flexibility and electricity price fluctuation.

ARC allows an incremental procedure to release operating reserves for energy dispatch. The ARC price is set by the higher of offered price or a Peaker Proxy Offer (Daily Peaker Proxy is calculated from monthly Peaker Proxy Heat Rate based on average CT

  • ffers in Real Time market during previous month) and

Daily Spot Price (Chicago spot gas) index.

slide-12
SLIDE 12

12

SP in MISO…(3)

Adequate Ramp Capability Pricing example: Start-up = $3,000 per start No-load = $750 per hour, Incremental energy = $100/MWh Capacity = 50 MW Total Single Hour Offer = ($3000/50) + ($750/50) + $100 = $175/MWh Total offer = $175/MWh Spot Gas = $7.44/MMBtu Peaker Proxy heat Rate = $175/$7.44 = 23.5 MMBtu /MWh The Proxy Peaker Price is determined as follow: On 6/1/2005 gas price = $6.32 Proxy Peaker Heat Rate = 23.5 Proxy Peaker Price = $6.32 * 23.5 = $148.5 Shortage Price is set higher of the offer or proxy peaker price. Step 2 uses additional capacity needed or longer period, 100% of the dispatch range. In step 2, the offer price released capacity segments is set to $1,000/MWh Interruptible generation is called on when MISO calls a Max Generation Emergency - Event Step 1 (NERC Engineering Emergency Alert 2) to Firm load shedding - Event Step 5 (NERC EEA 3).

slide-13
SLIDE 13

13

SP in NYISO

SP Trigger Reserves (10-min spinning, 10-min non- synchronized, 30-min ) by location shortage SP =

NYCA Eastern NY Long Island 10 m Spinning $500 $25 $25 10 m Total Reserve $150 $500 $25 30 m Reserve 200 MW @ $50 200 MW @ $100 200 MW @ $200 $25 $300

slide-14
SLIDE 14

14

SP in NYISO…(2)

NYCA Eastern NY Long Island 10 m Spinning

$850

= $500+$150+$200

$1400

=$500 + $25 + $$150 + $500 + $200+ $25

$1750

= $500 +$25+$25 + $150 + $500 + $25 +$200 + $25 + $300

10 m Total Reserve $350

= $150 + $200

$875

= $150 + $500 + $200 +$25

$1200

=$150 + $500 +$25 + $200 +$25 + $300 30 m Reserve

$200 $225

= $200 + $25

$525

= $300 + $200 + $25

slide-15
SLIDE 15

15

SP in NYISO…(3)

System Wide Regulation Demand Curve

  • 2. Energy Reserve co-optimization in Real Time

Guarantees that the clearing prices of energy, reserves, and regulation fully reflect opportunity cost Example: LMP = $120/MWh Offer Price = $50/MWh SP for LI = $70 + $1750/MWh

Regulation Demand Curve Value

Need > 25 MW to meet Target $300/MW Need < 25 MW to meet Target $250/MW

slide-16
SLIDE 16

16

SP in ISONE

SP Trigger via Locational Forward and Real Time Reserve Market that co-optimize Energy and Reserve

Physical Reserve Shortage = System not having enough capacity to meet the total energy and reserve requirement LMP = Marginal Unit’s Offer Price + Reserve Market Clearing Price Economic Reserve Shortage = System re- dispatch cost to procure reserve exceeds the value of Reserve Constraint Penalty Factors (System 10 Min spin - $50/MWh, System Total 10 Min Reserves $850/MWh, System 30 Min - $100/MWh, Local 30 Min - $50/MWh) LMP = Marginal Unit’s Offer Price + Re-dispatch cost to produce additional 1 MW of reserve

slide-17
SLIDE 17

17

SP in ERCOT

SP Trigger: Based on the supply demand conditions SP is based on the Australian model. At the beginning of the Resource Adequacy Cycle the

  • ffer cap is set to $1500/MWh (High Cap) but if the

Peaker Net Margin [the sum of all positive differences between the clearing price and estimated marginal cost of a hypothetical peaker with 10 Heat Rate] exceeds $175,000 per MW then the offer cap is reset to a lower level (Low Cap) to $500/MWh or $500 per MW per hr. for the remainder of that annual resource adequacy cycle.

slide-18
SLIDE 18

18

2005-6 Instances of SP Trigger

Year PJM MISO NYISO ISONE CAISO ERCOT 2005

July 26 – 5 h July 27 – 4.5 h None Aug 27 – 4 h None NA

2006

None In 2007 ARC June 15- 15m June 20 – 30 m July 18 – 9h July 19 – 9h Aug 1 – 5h Aug 2 – 11h Aug 3 – 6h Aug 1 – 40 m Aug 2 – 4 h 45 m NA In 2007 on April 3 – 7 fifteen minute intervals

slide-19
SLIDE 19

19

Methodology

  • Scarcity pricing defined as heat rates over 25,000

Btu/kWh. Only the most inefficient gas plants run at this level.

– For example, AES Redondo Beach in California is rated at or above 25,000 Btu/kWh in 2006. – This is not a magical heat rate, but just one that attempts to capture just the scarcity periods.

  • Hourly real time electricity prices from the RTO/ISOs

were compared to the daily day-ahead Gas Daily price.

– While scarcity can be approximated in day-ahead markets, the system events occur in real time. – Though gas prices may vary during the day, the daily price is a reasonable approximation of average variable fuel costs. – O&M costs were not factored into this analysis.

slide-20
SLIDE 20

20

Implied heat rates were calculated for every hour from 2004 to 2006

California ISO SP-15 Hourly Implied Heat Rates (2004 - 2006)

  • 20,000

20,000 40,000 60,000 80,000 100,000 120,000 5000 10000 15000 20000 25000 30000 Hourls Heat Rate (Btu/kWh)

Hours

slide-21
SLIDE 21

21

Implied heat rates over 25,000 Btu/kWh were considered scarcity

California ISO SP-15 Hours where Implied Heat Rate was Greater than 25,000 Btu/kWh (2004 - 2006) 25,000 35,000 45,000 55,000 65,000 75,000 85,000 95,000 105,000 50 100 150 200 250 300 Hours Btu/kWh

215 hours in SP-15 in 3 years were over 25

slide-22
SLIDE 22

22

Off-peak shortage conditions

  • Off-peak shortage can be related to issues with

ramping, scheduling (i.e. day-ahead forecast accuracy) and system events (e.g. outages) but not system capacity scarcity.

  • Off-peak shortage as a percentage of total
  • bserved scarcity ranged as follows for zones

and hubs covered in this analysis:

– ISONE: No instances – NYISO: 1% to 4% – PJM: 0% to 4% – MISO: 3% to 25% – CAISO: 25% to 36% – ERCOT: 4% to 6%

slide-23
SLIDE 23

23

Scarcity prices a function of missed load forecast

New York ISO Hudson Valley (Zone G) Hourly Real Time Price 5/30/06

$0 $100 $200 $300 $400 $500 $600 $700 $800 $900 $1,000 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 Hour Ending

Data source: Derived from NYISO real time hourly prices. http://www.nyiso.com/public/market_data/pricing_data.jsp Scarcity Pricing Hours

slide-24
SLIDE 24

24

Ramping can cause scarcity conditions

Data source: Derived from CAISO prices and Platts Gas Daily prices. http://oasis.caiso.com

California ISO Count of Scarcity Hours (2004-2006)

2 4 6 8 10 12 14 16 18 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 Hour Ending Count of Observations NP15

Ramping Hours

slide-25
SLIDE 25

25

As load and units come off in the evening, ramping scarcity exists

CAISO Systemwide Real Time Price 9/20/06

$- $50 $100 $150 $200 $250 $300 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 Hour Ending

As units come off from schedules, ramping scarcity causes spike.

Data source: Derived from CAISO prices. http://oasis.caiso.com

slide-26
SLIDE 26

26

As load and units come off in the evening, ramping scarcity exists

Data source: Derived from CAISO prices. http://oasis.caiso.com

SP15

  • 50

50 100 150 200 250 300 350 400

1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 Sum of HE01 Sum of HE02 Sum of HE03 Sum of HE04 Sum of HE05 Sum of HE06 Sum of HE07 Sum of HE08 Sum of HE09 Sum of HE10 Sum of HE11 Sum of HE12 Sum of HE13 Sum of HE14 Sum of HE15 Sum of HE16 Sum of HE17 Sum of HE18 Sum of HE19 Sum of HE20 Sum of HE21 Sum of HE22 Sum of HE23 Sum of HE24 9/20/06

SP15 OPR_DT Data RT_INTERVAL CNGS_ZONE

9/20/06 5-Minute Interval Prices

slide-27
SLIDE 27

27

Ramping causes scarcity prices to trigger in New York ISO

New York ISO Hudson Valley (Zone G) Hourly Real Time Price 10/18/06

$0 $50 $100 $150 $200 $250

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

Hour Ending

Morning ramping spike with heat rates exceeding 25,000 Btus/kWh

Data source: Derived from NYISO real time hourly prices. http://www.nyiso.com/public/market_data/pricing_data.jsp

slide-28
SLIDE 28

28

Scarcity pricing events in RTO/ISO markets from 2004 - 2006

Scarcity Events in RTO/ISO Markets

50 100 150 200 250 300 350 400 Hud Vly Long Island NYC West BGE NJ Hub West Hub NI Hub Mass Hub CT Zone FE Hub Cin Hub Ill Hub Mich Hub Minn Hub NP15 SP15 North South Northeast West Houston Hub Count of Observations 2004 2005 2006 ERCOT NYISO ISONE MISO CAISO PJM

Note: All electricity prices are from real time markets. PJM NI Hub did not start till May 2004. NYISO SMD2 market redesign occurred in Feb. 2005. MISO started market operation on April 2005. First Energy Hub (FE) did not start till Sep. 2005. CAISO software changes were implemented in 2006.

slide-29
SLIDE 29

29

Average heat rates during scarcity events differ by region

Average Heat Rate During Scarcity Events in RTO/ISO Markets

25,000 35,000 45,000 55,000 65,000 75,000 85,000 Hud Vly Long Island NYC West BGE NJ Hub West Hub NI Hub Mass Hub CT Zone FE Hub Cin Hub Ill Hub Mich Hub Minn Hub NP15 SP15 North South Northeast West Houston Hub

  • Avg. Heat Rate Btu/kWh

2004 2005 2006 ERCOT NYISO ISONE MISO CAISO PJM

Note: All electricity prices are from real time markets. PJM NI Hub did not start till May 2004. NYISO SMD2 market redesign occurred in Feb. 2005. MISO started market operation on April 2005. First Energy Hub (FE) did not start till Sep. 2005. CAISO software changes were implemented in 2006.

slide-30
SLIDE 30

30

Conclusion

  • Scarcity Pricing plays a different role based on

whether it is implemented in an energy only market or a capacity market

  • There can be shortage conditions even when

there is no “scarcity” in a traditional sense

  • Increased prices under scarcity conditions send

the right price signals to active Demand Response programs

  • Market manipulation under SP and Performance

metrics for generation investment should be addressed

  • Next step is to isolate off-peak hour shortage

conditions by shortage types