THE BENEFITS OF DYNAMIC PRICING IN MASS MARKETS Ahmad Faruqui, Ph. D.

advertisement
THE BENEFITS OF DYNAMIC PRICING IN
MASS MARKETS
Ahmad Faruqui, Ph. D.
Carnegie Mellon University
Pittsburgh, Pennsylvania
March 13, 2007
Précis
• Many utilities are considering the system-wide deployment of smart meters
►
►
They can improve utility operations and the cost savings can cover a substantial
portion of the multi-million dollar investment
However, depending on the utility, the “gap” between operational benefits and
AMI costs may still be quite large
• One way of bridging the gap is to use smart meters as a means of providing
“smart” prices to customers that would induce demand response, obviating
the need for expensive peaking capacity and energy
• As a bonus, smart pricing would eliminate an important inequity in existing
rate designs
►
Consumers who use relatively less power during expensive peak periods
subsidize others
2
The inequity in flat rates may amount to $4 billion dollars for
a state with 10 million customers
Load Shapes by Customer Type
(10 million customers total)
Consumption (kWh per Month)
600
500
Peak
Rates
Peak
Peak
400
Flat = $0.10/kWh
300
200
Peak = $0.20/kWh
Off-Peak
Off-Peak
Off-Peak
100
Off-Peak = $0.067/kWh
0
Flat
Average
Peaky
Amount of Cross-Subsidy
Per customer = $10/month
Total per month = $33.3 million
NPV (10 years) = $3.9 billion
3
Smart prices allow for risk sharing between suppliers
and consumers
Supplier
Risk
Flat Rate
Seasonal Rate
TOU
CPP-F
CPP-V
VPP
RTP
Consumer
Risk
4
Critical-peak pricing (CPP) is by far the most popular design
• It is essentially a time-of-use (TOU) rate on most days of the year
• When the power system encounters critical conditions, the peakperiod price rises to much higher but known levels, either on a dayahead or day-of basis
• In variable critical-peak pricing (VPP), the critical-peak price rises to
an unknown level that reflects actual market conditions
• Both of these rate designs approximate real-time pricing (RTP) rates
and are easier for mass market customers to deal with
5
A CPP rate will provide customers with substantial
opportunities to save money
Current Residential Rate vs. Cost-Based CPP/TOU All-In Rate
1.00
0.90
0.80
$0.90934/kWh
Current Rate
Critical Peak
New Rate
Rate ($/kWh)
0.70
0.60
0.50
0.40
0.30
0.20
0.10
$0.14807/kWh
$0.14824/kWh
$0.11312/kWh
0.00
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Hour of Day
6
For a vast majority of summer hours, the customer will save
money
Price Duration Curve
1.20
Rate ($/kWh)
1.00
Cost-Based CPP/TOU Rate
0.80
0.60
0.40
Current Rate
0.20
0.00
0
500
1000
1500
2000
Number of Hours in Summer Period
7
2500
Dynamic prices have had a substantial impact in a hot
climate such as California’s Central Valley
Figure 11
Hourly Load Shape - Complex Daily Share Model - Zone 4
2.5
kW Load
2.0
1.5
1.0
<-Peak Period->
0.5
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
Control 4
Treatment 4
8
They produce an impact even in a mild climate such as San
Francisco’s
Figure 8
Hourly Load Shape - Complex Daily Share Model - Zone 1
0.9
0.8
0.7
kW Load
0.6
0.5
0.4
0.3
<-Peak Period->
0.2
0.1
0.0
1
2
3
4
5
6
7
8
9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Hour
Control 1
Treatment 1
9
The higher the price, the greater is the drop in peak usage,
with the reduction varying with market segment
Residential Customer Response Curves
Decrease in Critical Peak Demand
0.0%
-5.0%
-10.0%
-15.0%
-20.0%
-25.0%
-30.0%
-35.0%
0.00
CAC w/ Technology
CAC
Average
Non-CAC
0.20
0.40
0.60
Critical Peak Rate ($/kWh)
10
0.80
1.00
Load reductions can be enhanced through enabling
technologies
Type of technology
60
50
40
30
20
10
0
Percent drop in critical peak load
Smart Meter
Smart
Thermostat
11
Gateway
Systems
Weighted
Average
Under traditional ratemaking, 50% of the customers would
be worse off under dynamic pricing
Distribution of Bill Impacts
20%
Electricity Bill Increase (Decrease)
15%
10%
5%
0%
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
-5%
-10%
Customers with Flatter Consumption
Customers with Peakier Consumption
-15%
Percentile of Customer Base
12
100%
That fear may keep customers from even trying out the new
rates
• And fear of that fear may keep us from even offering dynamic pricing
to customers, since we are anxious to “protect the customers from
themselves”
• How do we break out of this bubble?
13
Enter the risk premium
• Flat rates embody an implicit but very real risk premium that insures customers
against price volatility
• The risk premium is proportional to the volatility of loads, the volatility of spot prices
and the correlation between loads and spot prices
►
Thus, if load volatility is 0.2, price volatility is 0.6 and price-load correlation is 0.4, the
risk premium is about 5%
• π = exp( σL. σP. ρL,P )
Where:
►
►
►
►
π = Risk Premium
σL = Load Volatility
σP = Spot Price Volatility
ρL,P = Correlation Between Load and Spot Price
14
A Monte Carlo simulations suggest that a 3% risk premium
is a conservative estimate
Probability Distribution of Risk Premium
8%
3% risk premium
6%
5%
4%
3%
2%
1%
0%
0%
2%
4%
6%
8%
10%
12%
14%
16%
18%
20%
22%
24%
26%
28%
30%
32%
34%
36%
38%
40%
42%
44%
46%
48%
50%
52%
54%
56%
58%
60%
62%
64%
66%
68%
70%
Probability of Occuring
7%
Risk Premium
15
After crediting for the risk premium, dynamic pricing rates
become attractive for 70% of customers
Distribution of Bill Impacts
20%
Revenue Neutral
Electricity Bill Increase (Decrease)
15%
Risk Adjusted
10%
5%
0%
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
-5%
-10%
-15%
Customers with Flatter Consumption
Customers with Peakier Consumption
-20%
Percentile of Customer Base
16
100%
Enter demand response
• There is substantial evidence that dynamic pricing will lower critical
peak loads by more than 10% per average household
• The bigger the household’s monthly consumption level, the more will
be the load drop
• Customers in hot climate zones will exhibit the most demand
response
17
After crediting customers with the risk premium and
demand response, we can attract over 95% of customers
Distribution of Bill Impacts
20%
Revenue Neutral
Electricity Bill Increase (Decrease)
15%
Risk Adjusted
Load Shifting
10%
5%
0%
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
-5%
-10%
-15%
-20%
Customers with Flatter Consumption
Customers with Peakier Consumption
-25%
Percentile of Customer Base
18
100%
Aggregate MW impacts and financial benefits depend on the
number of participating customers
• The participation rate will depend on the deployment scenario and
marketing strategy
• A mandatory scenario will generate the highest number of
participants, followed by an opt-out scenario (around 70-90 percent)
and finally by an opt-in scenario (from 10 to 30 percent)
• In all cases, the CPP rate needs to generate substantial bill savings
for customers
19
Impacts vary by utility size and location
Impact on Four Representative U.S. Utilities
MW or Millions of 2007 Dollars
2,500
2,000
MW
$
1,500
MW
$
1,000
MW
500
$
MW
$
0
Large
Southeastern
Large
Western
Midsize
Mid-Atlantic
20
Small
Midwestern
Additional reading
• Brattle Group, The. “Quantifying the benefit of demand response for PJM,”
prepared for PJM Interconnection LLC. and MADRI, January 2007
• Faruqui, Ahmad. “Breaking out of the bubble: how dynamic pricing can
mitigate rate shock,” Public Utilities Fortnightly, March 2007.
• Federal Energy Regulatory Commission (FERC), The US. “Demand Response
and Advanced Metering,” Staff Report, August 2006
• North American Electric Reliability Corporation (NERC). “2006 Long-Term
Reliability Assessment,” October 16, 2006.
• Plexus Research, Inc., “Deciding on Smart Meters,” Edison Electric Institute,
September 2006.
21
Contact information
Ahmad Faruqui, Ph. D.
Principal
The Brattle Group
353 Sacramento Street, Suite 1140
San Francisco, CA 94111
Voice: 415.217.1026
Fax: 415.217.1099
Cell: 925.408.0149
Email: Ahmad.Faruqui@Brattle.Com
22
Download