ACC 2012 Tutorial http://accworkshop12.mit.edu The talk will review the many services needed in today's grid, and those that will be more important in the future. It will also review recent competitive equilibrium theory for the highly dynamic markets that may emerge in tomorrow's grid. In particular, to combat volatility from increasing penetration of renewable energy resources, there will be greater need for regulation services at various time-scales. There is enormous potential to secure these ancillary services via demand response. However, there is an obsession today with the promotion of real time prices to incentivize demand response. All evidence strongly suggests that this is a bad idea: 1) In 2011, massive price swings in the real-time market generated anger in Texas and New Zealand 2) Our own research shows that this is to be expected: in a completive equilibrium real-time prices will reach the choke up price (which was recently estimated at 1/4 million dollars). With transmission constraints, our research concludes that prices can go much higher. 3) A recent EIA study shows that consumers are scared of smart meters - they do not trust utility companies to experiment with their meters, or their power bills. We must then ask, is there any motivation to focus on markets in a real-time setting? The speaker believes there is none. Explanations will be given, and alternative visions will be proposed.

1. Markets for Differentiated Electric Power Products
Smart Grid Markets
Integration of Renewables, Pricing, Modeling, and Optimization
Emerging Topics in Interconnected Systems
Sean P. Meyn
Joint work with: In-Koo Cho, Anupama Kowli, Matias Negrete-Pincetic,
Ehsan Shafieeporfaard, Uday Shanbhag, and Gui Wang
Laboratory for Cognition & Control in Complex Systems
Department of Electrical and Computer Engineering
University of Florida
Thanks to NSF, AFOSR, and DOE / TCIPG
June 26, 2012

2. Markets for Differentiated Electric Power Products
Conclusions in advance
Traditional fossil fuels will be history to our great grandchildren
2 / 34

3. Markets for Differentiated Electric Power Products
Conclusions in advance
We need renewable energy, but how do we create a new energy
infrastructure to support this?
2 / 34

4. Markets for Differentiated Electric Power Products
Conclusions in advance
We need renewable energy, but how do we create a new energy
infrastructure to support this?
Some required elements:
2 / 34

5. Markets for Differentiated Electric Power Products
Conclusions in advance
We need renewable energy, but how do we create a new energy
infrastructure to support this?
Some required elements:
Electricity must treated as a service and not a commodity: Gas
turbine generation provides regulatory service. So could HVAC
Smart Grid programs have helped to create a framework for greater
service differentiation
Real time control will be an essential element to combat volatility and
ensure reliability
2 / 34

6. Markets for Differentiated Electric Power Products
Conclusions in advance
Traditional fossil fuels will be history to our great grandchildren
We need renewable energy, but how do we create a new energy
infrastructure to support this?
Some required elements:
Electricity must treated as a service and not a commodity: Gas
turbine generation provides regulatory service. So could HVAC
Smart Grid programs have helped to create a framework for greater
service differentiation
Real time control will be an essential element to combat volatility and
ensure reliability
Real time prices have little or no value here:
This is supported by theory and history.
2 / 34

7. Outline
1 Smart Grid in 2012
2 Some Science
3 Conclusions & Suggestions
4 References
3 / 34

8. Smart Grid in 2012
Nodal Power Prices
$20,000
per MWh Otahuhu
Stratford
$10,000
$0
0 4am 9am 2pm 7pm
Smart Grid 2012
4 / 34

9. Smart Grid in 2012
EIA 2011 Study
Smart grid legislative and regulatory policies and case studies
Many success stories:
Millions of smarter meters installed all over the globe
PNNL study: Automation of water heaters and other appliances
provided ancillary service in the Olympic peninsula
Large buildings such as hotels, and energy-intensive companies such
as IBM, Google, and ALCOA have contracts in place to help stabilize
the grid, encouraged by FERC Ruling 745∗
∗
Market-Based Demand Response Compensation Rule:
Electric utilities and retail market operators are now
required to pay demand response resources the market
price (LMP) for energy
5 / 34

10. Smart Grid in 2012
Increasing Leverage of Flexibility
Constellation Energy & NJP&L: Awards gift cards and rate reductions to residents
for control of air conditioners; company sells flexibility as ancillary service.
www.eia.gov/analysis/studies/electricity/pdf/smartggrid.pdf, December 12, 2011
Energy department to launch new energy innovation hub
focused on advanced batteries and energy storage.
www.energy.gov, February 7, 2012
Honeywell And Hawaiian Electric To Use Demand Response
To Integrate Renewables And Reduce Fossil Fuel Dependence.
www.honeywell.com, February 2, 2012
Axion Power’s PowerCube Battery Energy Storage
System Integrated Into PJM Utility Grid.
www.axionpower.com, November 22, 2011
First ’small-scale’ demand-side projects in PJM
providing frequency regulation.
www.sacbee.com, November 21, 2011
6 / 34

11. Smart Grid in 2012
EIA 2011 Study
Smart grid legislative and regulatory policies and case studies
Many success stories ... and failures
Residential consumers have high expectations,
Predictable cost savings
They may distrust those tampering with their appliances.
They distrust meters they believe interfere with appliances.
7 / 34

12. Smart Grid in 2012
EIA 2011 Study
Smart grid legislative and regulatory policies and case studies
Many success stories ... and failures
Residential consumers have high expectations,
Predictable cost savings
They may distrust those tampering with their appliances.
They distrust meters they believe interfere with appliances.
Moreover, the value of ancillary service obtained via demand response may
be reduced because of uncertainty of the level of consumer response.
7 / 34

13. Smart Grid in 2012
EIA 2011 Study
Smart grid legislative and regulatory policies and case studies
Many success stories ... and failures
Residential consumers have high expectations,
Predictable cost savings
They may distrust those tampering with their appliances.
They distrust meters they believe interfere with appliances.
Moreover, the value of ancillary service obtained via demand response may
be reduced because of uncertainty of the level of consumer response.
... yet, prices to devices are coming our way!∗
∗
Terry Boston, CEO PJM, ISGT 2012
7 / 34

14. Smart Grid in 2012
EIA 2011 Study
Smart grid legislative and regulatory policies and case studies
Moreover, the value of ancillary service obtained via demand response may
be reduced because of uncertainty of the level of consumer response.
... yet, prices to devices are coming our way!∗
∗
Terry Boston, CEO PJM, ISGT 2012
My concern: real-time pricing not TOU or contracts
7 / 34

15. Smart Grid in 2012
EIA 2011 Study
Case studies ... very little to say on real-time prices
"The active participation of final demand in the wholesale market is essential to managing
the greater intermittency of renewable resources and in limiting the ability of wholesale
electricity suppliers to exercise unilateral market power. A demand that is able to reduce its
consumption in real-time in response to higher prices limits the ability of suppliers
to exercise unilateral market power in a formal wholesale market such as the California ISO"
(http://www.stanford.edu/group/fwolak/cgi-
bin/sites/default/files/files/little_hoover_testimony_wolak_sept_2011.pdf) -F. Wolak
"Virtually all economists agree that the outcome [of the California crisis] was exacerbated by the inability of the demand side of the
market to respond to real or artificial supply shortages. This realization prompted my research stream on real-time electricity
pricing." - S. Borenstein
My concern: real-time pricing not TOU or contracts
8 / 34

16. Smart Grid in 2012
Winds Cause Price Spikes
Midwest ISO today: Friday afternoon, March 4, 2011 3:30 p.m.
-2000.00
9 / 34

17. Smart Grid in 2012
Winds Cause Price Spikes
Midwest ISO today: Friday afternoon, March 4, 2011 3:50 p.m.
-762.55
10 / 34

18. Smart Grid in 2012
Winds Cause Price Spikes
Midwest ISO today: Friday afternoon, March 4, 2011 4:15 p.m.
-1881.07
11 / 34

19. Smart Grid in 2012
Winds Cause Price Spikes
Midwest ISO today: Friday afternoon, March 4, 2011 4:30 p.m.
12 / 34

20. Smart Grid in 2012
Cold Causes Price Spikes
Texas today: Winter of 2011
Power Prices in Texas
$/MWh
3000
80
$/MWh January 31, 2011 February 2, 2011
2000
60
40
1000
20
10
0 0
−10
5am 10am 3pm 8pm
−500 5am 10am 3pm 8pm
13 / 34

21. Smart Grid in 2012
Cold Causes Price Spikes
Texas today: Winter of 2011
Power Prices in Texas
$/MWh
3000
80
$/MWh January 31, 2011 February 2, 2011
2000
60
40
1000
20
10
0 0
−10
5am 10am 3pm 8pm
−500 5am 10am 3pm 8pm
There will be multiple autopsies of the causes for the latest power breakdowns ... Who profited
off this near-meltdown and what can be done to incentivize power producers to maintain
adequate reserve capacity for emergencies rather than waiting for emergency windfalls? –
HOUSTON CHRONICLE, Feb 12, 2011
New report hits ERCOT, electricity deregulation: A report released Monday concludes that
electric deregulation has cost Texas residential consumers more than $11 billion in higher
rates... – Dallas Morning News, Feb 14, 2011
13 / 34

22. Smart Grid in 2012
Day-Ahead Market Outcomes
Texas today: Summer of 2011
ERCOT North Zone - August 1-30, 2011
Hourly day-ahead, daily on-peak, and monthly weighted average prices
3,000 hourly, day-ahead price
2,500
wholesale price ($/MWh)
daily, on-peak price
2,000 weighted average monthly
price ($188/MWh)
1,500
1,000
500
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31
Source: U.S. Energy Information Administration, based on the Electric Reliability Council of Texas (ERCOT).
Source: U.S. Energy Information Administration, based on the Electric Reliability Council of Texas (ERCOT).
Note: ERCOT North Zone includes Dallas/Fort Worth metrometro region and surrounding areas of Northeast Texas. On-Peak
Note: ERCOT North Zone includes Dallas/Fort Worth region and surrounding areas of Northeast Texas. On-Peak refers
to the 16-hour16-hour time block from hours ending to 10:00 p.m. 10:00 p.m. CDT on weekdays, excluding NERC holidays
refers to the time block from hours ending 7:00 a.m. 7:00 a.m. to CDT on weekdays, excluding NERC holidays.
14 / 34

23. Smart Grid in 2012
Madness in New Zealand
New Zealand today: March 25, 2011
A typical day in the New Zealand power market on the N. Island
Nodal Power Prices in NZ: $/MWh Otahuhu
Stratford
100
50
0
4am 9am 2pm 7pm http://www.electricityinfo.co.nz/
15 / 34

24. Smart Grid in 2012
Madness in New Zealand
New Zealand today: March 26, 2011
$25 million dollars extracted by the generators in just six hours
Nodal Power Prices in NZ: $/MWh Otahuhu
20,000 Stratford
10,000
0
4am 9am 2pm 7pm http://www.electricityinfo.co.nz/
16 / 34

25. Smart Grid in 2012
Madness in New Zealand
New Zealand today: March 26, 2011
>$20 million dollars demanded back from Genesis
Nodal Power Prices in NZ: $/MWh Otahuhu
20,000 Stratford
10,000
0
4am 9am 2pm 7pm http://www.electricityinfo.co.nz/
Preliminary view of NZ Electrical Authority: Genesis was not guilty of
“manipulative” ... or “deceptive” conduct. However, high prices threatened to
16 / 34

26. Smart Grid in 2012
Madness in New Zealand
New Zealand today: March 26, 2011
>$20 million dollars demanded back from Genesis
Nodal Power Prices in NZ: $/MWh Otahuhu
20,000 Stratford
10,000
0
4am 9am 2pm 7pm http://www.electricityinfo.co.nz/
Preliminary view of NZ Electrical Authority: Genesis was not guilty of
“manipulative” ... or “deceptive” conduct. However, high prices threatened to
undermine confidence in, and ... damage the integrity and reputation of the
wholesale electricity market 3:59 PM Friday May 6, 2011 www.nzherald.co.nz
16 / 34

27. Smart Grid in 2012
PNNL Prices to Devices Projects
Automation in the market Transactive Controls: Market-Based GridWiseTM Controls for Building Systems
$$$
Bid Price
P + kσ
P Mean Price
Bid Curve
Clearing Price
Current Zone
P – kσ
Desired or Temperature
Idea Set Point
Minimum Maximum
Set Point Set Point
Tset = 72oF Tmax = 77oF
Adjusted Zone
Tmin = 67oF
Tset,a = 70oF Tcurrent = 75oF
Comfort
Set Point Temperature
17 / 34

28. Smart Grid in 2012
PNNL Prices to Devices Projects
Automation in the market Transactive Controls: Market-Based GridWiseTM Controls for Building Systems
$$$
Bid Price
P + kσ
P Mean Price
Bid Curve
Clearing Price
Current Zone
P – kσ
Desired or Temperature
Idea Set Point
Minimum Maximum
Set Point Set Point
Tset = 72oF Tmax = 77oF
Adjusted Zone
Tmin = 67oF
Tset,a = 70oF Tcurrent = 75oF
Comfort
Set Point Temperature
Proportional control: Comfort = k × Price
17 / 34

29. Smart Grid in 2012
PNNL Prices to Devices Projects
Automation in the market Transactive Controls: Market-Based GridWiseTM Controls for Building Systems
$/MWh Mean Price
Zone Bid Price
Market Clearing Price
300
200
100
Hour
0
0 5 10 15 20 25
Proportional control: Comfort = k × Price
18 / 34

30. Smart Grid in 2012
PNNL Prices to Devices Projects
Automation in the market Transactive Controls: Market-Based GridWiseTM Controls for Building Systems
ConsumerAnger Mean Price
Zone Bid Price
Market Clearing Price
300
200
100
Hour
0
0 5 10 15 20 25
Proportional control: Comfort = k × Price
19 / 34

31. Smart Grid in 2012
MIT Prices to Devices Projects
Automation in the market Market-Based Control @ MIT
Fig. 4. Stochastic evolution of prices and demand Roozbehani et. al. 2010
$$$
Demand
Consumer Anger? 100
600
Ask Munzer et. al. Mean Price
Zone Bid Price
Demand
400
50
200
0
Hour
0 50 100 150
Real Time Prices Can Be Ugly
20 / 34

32. Smart Grid in 2012
MIT Prices to Devices Projects
Automation in the market Market-Based Control @ MIT
Fig. 4. Stochastic evolution of prices and demand Roozbehani et. al. 2010
$$$
Demand
Consumer Anger? 100
600
Ask Munzer et. al. Mean Price
Zone Bid Price
Demand
400
50
200
0
Hour
0 50 100 150
Real Time Prices Can Be Ugly Seen in Theory & Practice
20 / 34

33. Smart Grid in 2012
The Future as seen by FERC Today
FERC Order 755: RTMs on the microsecond scale
Regulation Required @ MISO
600
400
200
0
-200
-400
Time
New rules for fair treatment of resources participating in regulation markets
Current method of regulation compensation does not fairly account for the
regulation service provided.
21 / 34

34. Smart Grid in 2012
The Future as seen by FERC Today
FERC Order 755: RTMs on the microsecond scale
Regulation Required @ MISO
600
400
200
0
-200
-400
Time
New rules for fair treatment of resources participating in regulation markets
Current method of regulation compensation does not fairly account for the
regulation service provided.
Requires ISOs to pay resources based on actual service provided
21 / 34

35. Smart Grid in 2012
The Future as seen by FERC Today
FERC Order 755: RTMs on the microsecond scale
Regulation Required @ MISO
600
400
200
0
-200
-400
Time
New rules for fair treatment of resources participating in regulation markets
Current method of regulation compensation does not fairly account for the
regulation service provided.
Requires ISOs to pay resources based on actual service provided
Sounds fair enough!
21 / 34

36. Smart Grid in 2012
The Future as seen by FERC Today
FERC Order 755: RTMs on the microsecond scale
Regulation Required @ MISO
600
400
200
0
-200
-400
Time
New rules for fair treatment of resources participating in regulation markets
Possible payment plan, consider 1 of storage regulation,
T
d
Payment ∝ S(t)| dt
0 dt
22 / 34

37. Smart Grid in 2012
The Future as seen by FERC Today
FERC Order 755: RTMs on the microsecond scale
Regulation Required @ MISO
600
400
200
0
-200
-400
Time
New rules for fair treatment of resources participating in regulation markets
Possible payment plan, consider 1 of storage regulation,
T
d
Payment ∝ S(t)| dt
0 dt
Sounds game-able enough!
22 / 34

38. Some Science
Torque
Cost
Low
Speed
Some Science
23 / 34

39. Some Science
Torque
Cost
Low
Speed
Some Science
Concerning real-time pricing not TOU or contracts
23 / 34

40. Some Science
Equilibrium with Dynamics & Network Constraints
Entropic prices
Theorem 1: When dynamics (temporal constraints) are taken into
account, price is never equal to marginal cost [5, 4, 3, 1]
24 / 34

41. Some Science
Equilibrium with Dynamics & Network Constraints
Entropic prices
Theorem 1: When dynamics (temporal constraints) are taken into
account, price is never equal to marginal cost [5, 4, 3, 1]
Equilibrium price
The equilibrium price process is a function of equilibrium reserves:
P ∗ (t) = p∗ (Re (t))
The marginal value of power to the consumer.
24 / 34

42. Some Science
Equilibrium with Dynamics & Network Constraints
Entropic prices
Theorem 1: When dynamics (temporal constraints) are taken into
account, price is never equal to marginal cost [5, 4, 3, 1]
Equilibrium price
The equilibrium price process is a function of equilibrium reserves:
P ∗ (t) = p∗ (Re (t))
The marginal value of power to the consumer.
Proof: Lagrangian decomposition,
as in the static Second Welfare Theorem,
24 / 34

43. Some Science
Equilibrium with Dynamics & Network Constraints
Entropic prices
Theorem 1: When dynamics (temporal constraints) are taken into
account, price is never equal to marginal cost [5, 4, 3, 1]
Equilibrium price
The equilibrium price process is a function of equilibrium reserves:
P ∗ (t) = p∗ (Re (t))
The marginal value of power to the consumer.
Proof: Lagrangian decomposition,
as in the static Second Welfare Theorem,
or the proof of the Minimum Principle.
24 / 34

44. Some Science
Equilibrium with Dynamics & Network Constraints
Entropic prices
What is marginal value?
It is not always obvious. With the introduction of network constraints,
25 / 34

45. Some Science
Equilibrium with Dynamics & Network Constraints
Entropic prices
What is marginal value?
It is not always obvious. With the introduction of network constraints,
Prices can go well beyond marginal value (as defined in static model)
Prices can go well below zero
[Dynamic competitive equilibria in electricity markets. Wang et. al. 2011]
25 / 34

46. Some Science
Equilibrium with Dynamics & Network Constraints
Entropic prices
What is marginal value?
It is not always obvious. With the introduction of network constraints,
Prices can go well beyond marginal value (as defined in static model)
Prices can go well below zero
[Dynamic competitive equilibria in electricity markets. Wang et. al. 2011]
Without price-caps, Australia might look like an efficient equilibrium:
10,000 19,000 1,400
9,000
Victoria Demand 1,200 Tasmania 1,000
Price (Aus $/MWh)
8,000
Price (Aus $/MWh)
Volume (MW)
1,000
Volume (MW)
7,000
Demand
6,000 800
10,000
5,000 0
Prices 600
4,000
3,000 400
2,000 Prices
200 - 1,000
1,000 1,000
0 - 1,000 0 - 1,500
00:00
01:00
02:00
03:00
04:00
05:00
06:00
07:00
08:00
09:00
10:00
11:00
12:00
13:00
14:00
15:00
16:00
17:00
18:00
19:00
20:00
21:00
22:00
23:00
24:00
00:00
01:00
02:00
03:00
04:00
05:00
06:00
07:00
08:00
09:00
10:00
11:00
12:00
13:00
14:00
15:00
16:00
17:00
18:00
19:00
20:00
21:00
22:00
23:00
24:00
25 / 34

47. Some Science
Sustainable business?
Illinois: July 1998 California: July 2000
Spinning reserve prices PX prices $/MWh
5000 Purchase Price $/MWh 70
250
Previous week 60
4000 200
50
3000 150 40
2000 30
100
20
1000 50
10
0 0
Mon Tues Weds Thurs Fri Weds Thurs Fri Sat Sun Mon Tues Weds
Ontario: November, 2005 Texas: February 2, 2011
Marginal value of electricity,
Forecast Demand
Demand in MW Last Updated 11:00 AM Predispatch 1975.11 Dispatch 19683.5 $/MWh
21000 3000 Average price
is usually $30
18000
2000
$250,000/MWh (?)
15000
Hourly Ontario Energy Price $/MWh Last Updated 11:00 AM Predispatch 72.79 Dispatch 90.82
2000 1000
Forecast Prices
1500
$/MWh
1000
500 0
0 3 6 9 12 15 18 21 3 6 9 12 15 18 21 3 6 9 12 15 18 21 Time −500 5am 10am 3pm 8pm
Tues Weds Thurs
26 / 34

48. Some Science
Sustainable business?
Illinois: July 1998 California: July 2000
Spinning reserve prices PX prices $/MWh
5000 Purchase Price $/MWh 70
250
Previous week 60
4000 200
50
3000 150 40
2000 30
100
20
1000 50
10
0 0
Mon Tues Weds Thurs Fri Weds Thurs Fri Sat Sun Mon Tues Weds
Ontario: November, 2005 Texas: February 2, 2011
Marginal value of electricity,
Forecast Demand
Demand in MW Last Updated 11:00 AM Predispatch 1975.11 Dispatch 19683.5 $/MWh
21000 3000 Average price
is usually $30
18000
2000
$250,000/MWh (?)
15000
Hourly Ontario Energy Price $/MWh Last Updated 11:00 AM Predispatch 72.79 Dispatch 90.82
2000 1000
Forecast Prices
1500
$/MWh
1000
500 0
0 3 6 9 12 15 18 21 3 6 9 12 15 18 21 3 6 9 12 15 18 21 Time −500 5am 10am 3pm 8pm
Tues Weds Thurs
However,
26 / 34

49. Some Science
Sustainable business?
Illinois: July 1998 California: July 2000
Spinning reserve prices PX prices $/MWh
5000 Purchase Price $/MWh 70
250
Previous week 60
4000 200
50
3000 150 40
2000 30
100
20
1000 50
10
0 0
Mon Tues Weds Thurs Fri Weds Thurs Fri Sat Sun Mon Tues Weds
Ontario: November, 2005 Texas: February 2, 2011
Marginal value of electricity,
Forecast Demand
Demand in MW Last Updated 11:00 AM Predispatch 1975.11 Dispatch 19683.5 $/MWh
21000 3000 Average price
is usually $30
18000
2000
$250,000/MWh (?)
15000
Hourly Ontario Energy Price $/MWh Last Updated 11:00 AM Predispatch 72.79 Dispatch 90.82
2000 1000
Forecast Prices
1500
$/MWh
1000
500 0
0 3 6 9 12 15 18 21 3 6 9 12 15 18 21 3 6 9 12 15 18 21 Time −500 5am 10am 3pm 8pm
Tues Weds Thurs
However,
Theorem 2: In this equilibrium, the average price is
precisely the average marginal cost
Proof: Lagrangian relaxation of initial condition.
26 / 34

50. Some Science
Sustainable business?
Illinois: July 1998 California: July 2000
Spinning reserve prices PX prices $/MWh
5000 Purchase Price $/MWh 70
250
Previous week 60
4000 200
50
3000 150 40
2000 30
100
20
1000 50
10
0 0
Mon Tues Weds Thurs Fri Weds Thurs Fri Sat Sun Mon Tues Weds
Ontario: November, 2005 Texas: February 2, 2011
Marginal value of electricity,
Forecast Demand
Demand in MW Last Updated 11:00 AM Predispatch 1975.11 Dispatch 19683.5 $/MWh
21000 3000 Average price
is usually $30
18000
2000
$250,000/MWh (?)
15000
Hourly Ontario Energy Price $/MWh Last Updated 11:00 AM Predispatch 72.79 Dispatch 90.82
2000 1000
Forecast Prices
1500
$/MWh
1000
500 0
0 3 6 9 12 15 18 21 3 6 9 12 15 18 21 3 6 9 12 15 18 21 Time −500 5am 10am 3pm 8pm
Tues Weds Thurs
However,
Theorem 2: In this equilibrium, the average price is
precisely the average marginal cost
Proof: Lagrangian relaxation of initial condition.
Is this a sustainable business?
26 / 34

51. Some Science
More Engineering: Where is the Missing Money?
Addressing FERC Order 755
27 / 34

52. Some Science
More Engineering: Where is the Missing Money?
Addressing FERC Order 755
World-view, from eyes of a coal generator operator:
Control must respect dynamics & costs
27 / 34

53. Some Science
More Engineering: Where is the Missing Money?
Addressing FERC Order 755
World-view, from eyes of a coal generator operator:
Control must respect dynamics & costs
Cost of generation:
Includes G and dt G,
d
shut-down, O&M, investment, ...
27 / 34

54. Some Science
More Engineering: Where is the Missing Money?
Addressing FERC Order 755
World-view, from eyes of a coal generator operator:
Control must respect dynamics & costs
Cost of generation:
Torque
Includes G and dt G,
d
Cost
Low
shut-down, O&M, investment, ...
What is “marginal cost”?
Speed
27 / 34

55. Some Science
More Engineering: Where is the Missing Money?
Addressing FERC Order 755
World-view, from eyes of a coal generator operator:
Control must respect dynamics & costs
Cost of generation:
Torque
Includes G and dt G,
d
Cost
Low
shut-down, O&M, investment, ...
What is “marginal cost”?
Speed
Theorem 3: If c(G, dt G)
d
= ce (G) + cw ( dt G)
d
then
27 / 34

56. Some Science
More Engineering: Where is the Missing Money?
Addressing FERC Order 755
World-view, from eyes of a coal generator operator:
Control must respect dynamics & costs
Cost of generation:
Torque
Includes G and dt G,
d
Cost
Low
shut-down, O&M, investment, ...
What is “marginal cost”?
Speed
Theorem 3: If c(G, dt G)
d
= ce (G) + cw ( dt G)
d
then
E[P ∗ ] = E[ ce (G)]
27 / 34

57. Some Science
More Engineering: Where is the Missing Money?
Addressing FERC Order 755
World-view, from eyes of a coal generator operator:
Control must respect dynamics & costs
Cost of generation:
Torque
Includes G and dt G,
d
Cost
Low
shut-down, O&M, investment, ...
What is “marginal cost”?
Speed
Theorem 3: If c(G, dt G)
d
= ce (G) + cw ( dt G)
d
then
E[P ∗ ] = E[ ce (G)] =⇒ lots of missing money
27 / 34

58. Some Science
More Engineering: Where is the Missing Money?
Addressing FERC Order 755
World-view, from eyes of a coal generator operator:
Control must respect dynamics & costs
Cost of generation:
Torque
Includes G and dt G,
d
Cost
Low
shut-down, O&M, investment, ...
What is “marginal cost”?
Speed
Theorem 3: If c(G, dt G)
d
= ce (G) + cw ( dt G)
d
then
E[P ∗ ] = E[ ce (G)] =⇒ lots of missing money
Competitive equilibrium never compensates for “wear and tear”.
27 / 34

59. Conclusions & Suggestions
Conclusions & Suggestions
28 / 34

60. Conclusions & Suggestions
Economics and Engineering
Arbitrage – between Nukes and Turbines?
29 / 34

61. Conclusions & Suggestions
Economics and Engineering
Arbitrage – between Nukes and Turbines?
Any power engineer knows: Nuclear generation and gas turbines provide
different services, and are not just delivering electrons.
29 / 34

62. Conclusions & Suggestions
Economics and Engineering
Arbitrage – between Nukes and Turbines?
Any power engineer knows: Nuclear generation and gas turbines provide
different services, and are not just delivering electrons.
So FERC is on the right track ...
29 / 34

63. Conclusions & Suggestions
Economics and Engineering
Arbitrage – between Nukes and Turbines?
Any power engineer knows: Nuclear generation and gas turbines provide
different services, and are not just delivering electrons.
So FERC is on the right track ...
T
d
What about the other policy makers? Payment ∝ S(t)| dt
0 dt
29 / 34

64. Conclusions & Suggestions
Economics and Engineering
Arbitrage – between Nukes and Turbines?
Any power engineer knows: Nuclear generation and gas turbines provide
different services, and are not just delivering electrons.
So FERC is on the right track ...
T
d
What about the other policy makers? Payment ∝ S(t)| dt
0 dt
“ ... One result of this divorce of the theory from its subject
matter has been that the entitites whose decisions economists
are engaged in analyzing have not been made he subject of study
and in consequence lack any substance. ...consumers without
humanity, firms without organization, and even exchange without
markets”
R. Coase, 1988
29 / 34

65. Conclusions & Suggestions
Summary of Research at Illinois & Florida
The current RTM paradigm must be reconsidered
Conclusions of our research, under the most optimistic assumptions
Volatile prices are to be expected in real time electricity markets in an
efficient equilibrium.
30 / 34

66. Conclusions & Suggestions
Summary of Research at Illinois & Florida
The current RTM paradigm must be reconsidered
Conclusions of our research, under the most optimistic assumptions
Volatile prices are to be expected in real time electricity markets in an
efficient equilibrium. Prices are never equal to marginal cost.
30 / 34

67. Conclusions & Suggestions
Summary of Research at Illinois & Florida
The current RTM paradigm must be reconsidered
Conclusions of our research, under the most optimistic assumptions
Volatile prices are to be expected in real time electricity markets in an
efficient equilibrium. Prices are never equal to marginal cost.
Risk to suppliers: Average prices coincide with average marginal cost.
30 / 34

68. Conclusions & Suggestions
Summary of Research at Illinois & Florida
The current RTM paradigm must be reconsidered
Conclusions of our research, under the most optimistic assumptions
Volatile prices are to be expected in real time electricity markets in an
efficient equilibrium. Prices are never equal to marginal cost.
Risk to suppliers: Average prices coincide with average marginal cost.
Why stay in such a risky business?
30 / 34

69. Conclusions & Suggestions
Summary of Research at Illinois & Florida
The current RTM paradigm must be reconsidered
Conclusions of our research, under the most optimistic assumptions
Volatile prices are to be expected in real time electricity markets in an
efficient equilibrium. Prices are never equal to marginal cost.
Risk to suppliers: Average prices coincide with average marginal cost.
Why stay in such a risky business?
The real world
30 / 34

70. Conclusions & Suggestions
Summary of Research at Illinois & Florida
The current RTM paradigm must be reconsidered
Conclusions of our research, under the most optimistic assumptions
Volatile prices are to be expected in real time electricity markets in an
efficient equilibrium. Prices are never equal to marginal cost.
Risk to suppliers: Average prices coincide with average marginal cost.
Why stay in such a risky business?
The real world
Volatile prices are observed all over the world
Benefits to consumers are not clear, and innovation is slow
30 / 34

71. Conclusions & Suggestions
Summary of Research at Illinois & Florida
The current RTM paradigm must be reconsidered
Conclusions of our research, under the most optimistic assumptions
Volatile prices are to be expected in real time electricity markets in an
efficient equilibrium. Prices are never equal to marginal cost.
Risk to suppliers: Average prices coincide with average marginal cost.
Why stay in such a risky business?
The real world
Volatile prices are observed all over the world
Benefits to consumers are not clear, and innovation is slow
Market power is a reality, and symmetric information is absurd:
Strategic behavior can lead to a new crisis each year!
30 / 34

72. Conclusions & Suggestions
The current RTM paradigm must be reconsidered
Why have real time markets?
No value has been demonstrated for real-time markets.
31 / 34

73. Conclusions & Suggestions
The current RTM paradigm must be reconsidered
Why have real time markets?
No value has been demonstrated for real-time markets.
Empirical evidence: We cannot distinguish robbery from efficiency.
31 / 34

74. Conclusions & Suggestions
The current RTM paradigm must be reconsidered
Why have real time markets?
No value has been demonstrated for real-time markets.
Empirical evidence: We cannot distinguish robbery from efficiency.
Is This A Free Market For Fire Fighters?
31 / 34

75. Conclusions & Suggestions
The current RTM paradigm must be reconsidered
Why have real time markets?
No value has been demonstrated for real-time markets.
Empirical evidence: We cannot distinguish robbery from efficiency.
Is This A Free Market For Fire Fighters?
Why then would you use real-time prices to control devices?
31 / 34

76. Conclusions & Suggestions
The current RTM paradigm must be reconsidered
Why have real time markets?
No value has been demonstrated for real-time markets.
Empirical evidence: We cannot distinguish robbery from efficiency.
Is This A Free Market For Fire Fighters?
Why then would you use real-time prices to control devices?
The EIA study shows that there are alternatives
31 / 34

77. Conclusions & Suggestions
The current RTM paradigm must be reconsidered
Alternatives
TOU prices for peak shaving, and
Contracts for real-time demand-response services
32 / 34

78. Conclusions & Suggestions
The current RTM paradigm must be reconsidered
Alternatives
TOU prices for peak shaving, and
Contracts for real-time demand-response services
Successful contracts today: Constellation Energy, Alcoa, residential pool
pumps, commercial buildings (see new work at Univ. of Florida), ...
Efficiency loss, but utility and consumers each have reliable services
32 / 34

79. Conclusions & Suggestions
The current RTM paradigm must be reconsidered
Alternatives
New Smart appliances that can facilitate these contracts
Control theory to make this all work:
We don’t know why the grid is so robust today. Introducing
all of these dynamics will lead to new control challenges.
32 / 34

80. Conclusions & Suggestions
The current RTM paradigm must be reconsidered
Alternatives
New Smart appliances that can facilitate these contracts
Control theory to make this all work:
We don’t know why the grid is so robust today. Introducing
all of these dynamics will lead to new control challenges.
Energy policy that is guided by an understanding of both physics and
economics
32 / 34

81. Conclusions & Suggestions
The current RTM paradigm must be reconsidered
Alternatives
New Smart appliances that can facilitate these contracts
Control theory to make this all work:
We don’t know why the grid is so robust today. Introducing
all of these dynamics will lead to new control challenges.
Energy policy that is guided by an understanding of both physics and
economics
Thank You!
32 / 34

82. Conclusions & Suggestions
Pre-publication version for on-line viewing. Monograph available for purchase at your favorite retailer August 2008 Pre-publication version for on-line viewing. Monograph to appear Februrary 2009
More information available at http://www.cambridge.org/us/catalogue/catalogue.asp?isbn=9780521884419
Control Techniques Markov Chains
FOR and
Complex Networks Stochastic Stability
P n (x, · ) − π f →0
sup Ex [SτC (f )] < ∞
C
π(f ) < ∞
∆V (x) ≤ −f (x) + bIC (x)
Sean Meyn S. P. Meyn and R. L. Tweedie
References
33 / 34

83. References
References
G. Wang, M. Negrete-Pincetic, A. Kowli, E. Shafieepoorfard, S. Meyn, and U. Shanbhag.
Dynamic competitive equilibria in electricity markets. In A. Chakrabortty and M. Illic,
editors, Control and Optimization Theory for Electric Smart Grids. Springer, 2011.
M. Negrete-Pincetic and S. Meyn. Where is the Missing Money? The impact of generation
ramping costs in electricity markets. In preparation, 2012.
G. Wang, A. Kowli, M. Negrete-Pincetic, E. Shafieepoorfard, and S. Meyn.
A control theorist’s perspective on dynamic competitive equilibria in electricity markets. In
Proc. 18th World Congress of the International Federation of Automatic Control (IFAC),
Milano, Italy, 2011.
S. Meyn, M. Negrete-Pincetic, G. Wang, A. Kowli, and E. Shafieepoorfard. The value of
volatile resources in electricity markets. In Proc. of the 10th IEEE Conf. on Dec. and
Control, Atlanta, GA, 2010.
I.-K. Cho and S. P. Meyn. Efficiency and marginal cost pricing in dynamic competitive
markets with friction. Theoretical Economics, 5(2):215–239, 2010.
H. Hao, A. Kowli, T. Middelkoop, P. Barooah, and S. Meyn. Using flexible HVAC power
consumption of commercial buildings for regulation service to the grid. UF TR, 2012
U.S. Energy Information Administration. Smart grid legislative and regulatory policies and
case studies. December 12 2011.
http://www.eia.gov/analysis/studies/electricity/pdf/smartggrid.pdf
34 / 34