Embed Size (px)
Citation preview
Residen'al Transac've Control Field Results ISGT, Washington, DC 19 February 2014
Steve Widergren Principal Engineer
1 PNNL-SA-XXXX
gridSMART® RTPda Demo
! First real-time market at distribution feeder level with a tariff approved by the PUC of Ohio
! Value streams ! Energy purchase benefit: function of PJM LMP ! Capacity benefits: distribution feeder and system
gen/trans limitations, e.g., peak shaving ! Ancillary services benefits: characterized, but not
part of the tariff ! Uses market bidding mechanism to perform
distributed optimization – transactive energy ! ~200 homes bidding on 4 feeders ! Separate market run on each feeder ! “Double auction” with 5 minute clearing
! HVAC automated bidding ! Smart thermostat and home energy manager ! Homeowner sets comfort/economy preference ! Can view real-time and historical prices to make
personal choices 2
RTP System
3
Home
Dispatch System
Service Provider
Opera'ons
Residen'al Energy Mgmt
System
Consumer Display
Programmable Thermostat
Meter usage
supply, usage information
bids
~200 homes on 4 feeders
clearing price, usage
bids, clearing
price
Opera&ons Center Field
Wholesale Market 5 minute nodal energy prices
monthly bill
Pbase
RTP Market Uncongested Condi'ons
Retail RTP based on wholesale real-‐3me LMP (Base RTP)
Unresponsive Loads
Q, Load (MW)
P, P
rice
($/M
Wh)
Responsive Loads
Demand Curve: sorted (P, Q) bids from RTPDA customers
Pclear =
Qclear
Feeder Capacity
Varies every 5-‐min
Feeder Supply Curve
! Market clears every 5-min (~match AC load cycle)
! When uncongested:
! Quantity (Qclear) varies with demand curve
! Price (Pclear) is constant, equal to Base RTP
Qmin Qmax
Market clears at intersec3on of supply & demand curves
4
RTP Market Conges'on Condi'ons
5
Analysis 6
• RTPda analysis report draR in review • Period of study: 1 Jun – 30 Sep 2013 • Topics covered
– System impacts • Wholesale purchases • Capacity: system and feeder peak management • Spinning reserve poten'al
– Household impacts • Bills • Thermostat sta's'cs • HVAC energy use compared with quan'ty bid
– Load sensi'vity to price • Some items not covered here
– Customer sa'sfac'on – Revenue impacts
Wholesale Purchase Impacts 7
Over-‐cooling Observa'on 8
Observed indoor temperature driven below temperature set point due to frequent congestion events making normal prices appear inexpensive
Wholesale Purchase Impacts
• Raw Results – Wholesale energy use +1.9% and cost +.7% higher than control group
• Temperature Corrected Results – Wholesale energy use -‐5.3% and cost -‐5% lower than control group
• Simula&on Results – Wholesale energy use -‐1.2% and cost -‐2.5% lower than control group
9
RTP Load Response to Conges'on -‐ Actual
10
−4 −2 0 2 4 6 8−30
−20
−10
0
10
20
30
40Aggregate response impact hot and peak
Hour
%
RTPda(2h) − ControlRTPda(4h) − ControlTest start2h test end4h test end
70 2-hour events (blue) were held over a wide range (peak v off-peak energy use and mild v hot temperatures 26 4-hour events (red) had greater proportion of hot – peak periods when more HVAC resources are available to curtail
% R
TP L
oad
System & Feeder Capacity Impacts -‐ Simulated 11
System Congestion Control (blue) v 25% RTP
Dem
and
(kW
)
Dem
and
(kW
)
Feeder Congestion Control (blue) v 25% RTP
Simulation calibrated from field data: System peaks longer than feeder peaks Results: 50% RTP household penetration implies
~ 6.5% peak system load reduction, ~10% peak feeder load reduction
% P
enet
ratio
n R
TP h
ouse
s
% P
enet
ratio
n R
TP h
ouse
s % Feeder Peak Reduction % System Peak Reduction
Swarm of Responses to Price 12
Actual load response versus LMP for about 12,000 5-min data points covering the period June–September 2013 Bottom shows histogram of the frequency of LMPs up to $100/MWh
Diff
eren
ce in
Loa
d be
twee
n R
TP a
nd n
on-R
TP
Con
trol
Gro
up (k
W/h
ouse
)
The trend line illustrates the
systemic response of lower load to
higher price
Sensitivity of Load to Price
Temperature Sensi'vity of Response -‐ Simulated
13
Change in load versus relative price (in standard deviations from average price) between control and RTP groups broken into outdoor air temperature bins. Period is June-September 2013, effects of congestion events not modeled
Reduce load to price increase
Increase load to price
increase likely due to device “recovery”
Acknowledgement & Disclaimer Team : K Subbarao, JC Fuller, DP Chassin, C Marinovici, A Somani, JL Hammerstrom, S Widergren Acknowledgment: This work is supported in part by the Pacific Northwest Na'onal Laboratory operated for the U.S. Department of Energy by Bagelle under Contract DE-‐AC65-‐76RLO1830. Disclaimer: This report was prepared as an account of work sponsored by an agency of the United States Government. Neither the United States Government nor any agency thereof, nor any of their employees, makes any warranty, express or implied, or assumes any legal liability or responsibility for the accuracy, completeness, or usefulness of any informa'on, apparatus, product, or process disclosed, or represents that its use would not infringe privately owned rights. Reference herein to any specific commercial product, process, or service by trade name, trademark, manufacturer, or otherwise does not necessarily cons'tute or imply its endorsement, recommenda'on, or favoring by the United States Government or any agency thereof. The views and opinions of authors expressed herein do not necessarily state or reflect those of the United States Government or any agency thereof.
16
