POWER SYSTEM PLANNING WITH DISTRIBUTED ENERGY RESOURCES
NEW TRADEOFFS, METHODS AND INSIGHTS
Jesse D. Jenkins EPRI 36th Seminar on Fuels, Power Markets, and Resource Planning November 8th, 2017 -- Washington, DC
17cv01906 Sierra Club v. EPA
ED_001523_00006099-00001
17cv01906 Sierra Club v. EPA
ED_001523_00006099-00002
17cv01906 Sierra Club v. EPA
ED_001523_00006099-00003
17cv01906 Sierra Club v. EPA
ED_001523_00006099-00004
17cv01906 Sierra Club v. EPA
ED_001523_00006099-00005
Distribution of 2015 annual average nodal LMPs in PJM More than three quarters of nodes between $21-40/MWh
50.4%
26.2%
Approximately 3 percent of nodes with very high locational value, 3 10 times the average
0.1%
0.1%
8.4%
7.3%
2.3%
0.9%
0.4%
0.4%
0.5%
2.9% .
<1 1-10 11-20 21-30 31-40 41-50 51-60 61-70 71-80 81-90 91-100 >100
USD per MWh
See: Perez-Arriaga et al. 2016. The Utility of the Future: an MIT Energy Initiative response to an industry in transition.
Cambridge, MA: Massachusetts Institute of Technology, http://erergy.mit.edu/research/utility-future-study/
6
17cv01906 Sierra Club v. EPA
ED_001523_00006099-00006
Network capacity benefits of distributed solar PV in California
(Cohen, Kauzmann & Callaway, 2016)
$10-60 oer
>$60 per kW per year
Year 2025
Interc by Fe<
` 'V
0.01 -0,20 0,21 -0,50 0,51 -1,00 1.01 -2.00 2.01 - 5,00 5.01 +
See: M.A. Cohen, P.A. Kauzmann, D.S. Callaway, Effects of diSributed PV generation on California's distribution system,
part 2: Economic analysis, Solar Energy, Volume 128, 2016, 139-152
7
17cv01906 Sierra Club v. EPA
ED_001523_00006099-00007
Marginal value of distribution network losses avoided by distributed solar PV as penetration increases (Texas ERCOT Example)
25%
3% Avg Distribution Losses
9% Avg Distribution Losses
CD s O)
a Q 20% -
0 >
0
o
15%
05
Locational value "premium" from distribution loss avoidance may be 6-19% of average wholesale LMP for the first few PV systems installed, but falls steadily as PV penetration increases.
o o
w _CWoeoDi
10%
o
5%
0% 0%
5%
10%
15%
20%
25%
30%
35%
Penetration Level as Percent of Annual Energy from Distributed Solar PV
See: Perez-Arriaga et al. 2016. The Utility of the Future: an MIT Energy Initiative response to an industry in transition. Cambridge, MA: Massachusetts Institute of Technology, http://erergy.mit.edu/research/utility-future-study/
17cv01906 Sierra Club v. EPA
ED_001523_00006099-00008
Distribution of load curtailment/self-generation necessary to accommodate load growth without network reinforcement - European urban network case
. adi
\ a ? I > d 1 ja d >>v, d
Aiu:
L. i.! t : -aa :h
I' I.f ..<1 i >i I>,*. d
L<-\rl of CuruihiH'in ( ri'imUoii : MW j . ... .
* f M M''
Source: Jenkins, Luke & Vargara, forthcoming
4 *
17cv01906 Sierra Club v. EPA
9
ED_001523_00006099-00009
System Load (MW)
Load duration curve for ISO New England, 2011-2015, all hours.
28,000
26,000 24,000 22,000 20,000
Accommodating each marginal increment of load growth without upgrades requires both more MWs and more hours of net load reduction.
18,000
16,000
14,000
12,000
10,000
8,000 0%
10% 20% 2011
30% 40% 50%
2012
2013
60% 70% 80% 2014 -.......2015
90% 100%
Source: ISO New England (2015), "ISO New England's Internal Market Monitor 2015 Annual Markets Report."
17cv01906 Sierra Club v. EPA
ED_001523_00006099-00010
System Load (MW)
Load duration curve for ISO New England, 2011-2015, top 5% hous
28,000
26,000
24,000
22,000
20,000
18,000 16,000 14,000 12,000 10,000
"Peakiest" load hours may be curtailed by price responsive or flexible demand:
5% decline in peak demand can be achieved via curtailment during only -20-40 hours of the year.
10% decline can be achieved with -50-100 hours.
8,000 . -........................................
-............. --r
0%
1%
2%
3%
5%
2011
2012
2013
2014
2015
Source: ISO New England (2015), "ISO New England's Internal Market Monitor 2015 Annual Markets Report."
17cv01906 Sierra Club v. EPA
ED_001523_00006099-00011
Economies of unit scale vs locational value
Utility Scale
C&l Scale
Residential Scale
17cv01906 Sierra Club v. EPA
ED_001523_00006099-00012
Capital annuity and fixed O&M ($1,000/MW-yr)
$400 $350 $300 $250 $200 $150 $100
$50 $0
Estimated economies of unit scale for fixed-tilt U.S. solar PV systems: Annual cost of ownership in 2015 and projected for 2025
2015
(high cost estimate)
2025 (medium cost estimate)
(low cost estimate)
See: Perez-Arriaga et al. 2016. The Utility of the Future: an MIT Energy Initiative response to an industry in transition. Cambridge, MA: Massachusetts Institute of Technology, http://energy.mit.edu/research/utility-future-study/
17cv01906 Sierra Club v. EPA
ED_001523_00006099-00013
Estimated economies of unit scale for Li-ion energy storage systems (1:2 power to energy ratio): Annual cost of ownership in 2015 and projected for 2025
$600 T
2015
(high cost estimate)
2025 (medium cost estimate)
(low cost estimate)
$500
See: Perez-Arriaga et al. 2016. The Utility of the Future: an MIT Energy Initiative response to an industry in transition. Cambridge, MA: Massachusetts Institute of Technology, http://energy.mit.edu/research/utility-future-study/
17cv01906 Sierra Club v. EPA
ED_001523_00006099-00014
I- _____ J
Comparison of locational value and incremental unit costs for solar PV systems: Long Island, New York example, "high value" case
Average value per MWhproduced
$160 $140 $120
Distributed opportunity costs
148.7
$100 $80
41.2
2.9
$60
11.1
$40
5.6
24.0
$20
0.0
84.7
58.4
$0
Locational
Locational Conservation Network
energy value: energy value:
voltage
investment
transmission distribution
reduction
deferral
losses
Locational values
Generation capacity premium
Reliability Total locational 1-2 MW systeml-10 kW system value
Incremental unit costs relative to 10-100 MW system
See: Perez-Arriaga et al. 2016. The Utility of the Future: an MIT Energy Initiative response to an industry in transition. Cambridge, MA:
Massachusetts Institute of Technology, http://energy.mit.edu/research/utility-future-study/
15
17cv01906 Sierra Club v. EPA
ED_001523_00006099-00015
Comparison of locational value and incremental unit costs for solar PV systems: Mohawk Valley, New York example, "average value" case
$180
$160
158.6
Average value per MWhproduced
$140
$120
$100
$80
$60
$40
$20
$0
Locational
Locational Conservation Network
energy value: energy value:
voltage
investment
transmission distribution
reduction
deferral
losses
Generation capacity premium
Locational values
Reliability Total locational 1-2 MW system1-10 kW system value
Incremental unit costs relative to 10-100 MW system
See: Perez-Arriaga et al. 2016. The Utility of the Future: an MIT Energy Initiative response to an industry in transition. Cambridge, MA:
Massachusetts Institute of Technology, http://energy.mit.edu/research/utility-future-study/
16
17cv01906 Sierra Club v. EPA
ED_001523_00006099-00016
17cv01906 Sierra Club v. EPA
ED_O01523_00006099-00017
17cv01906 Sierra Club v. EPA
Hours
18
ED_001523_00006099-00018
17cv01906 Sierra Club v. EPA
ED_O01523_00006099-00019
Load Duration and Screening Curve Based Methods
17cv01906 Sierra Club v. EPA
20
ED_001523_00006099-00020
I
Minimize: (1) Sum of investment, fixed O&M and network reinforcement costs; (2) Sum of variable O&M and fuel costs per hour; (3) Sum of cost (or value) of other services provided by DERs
(e.g. voltage regulation, back-up power); (4) Sum of cost of curtailed demand and unmet operating reserves for each
hour
Subject to... Operational constraints on thermal generators, dispatchable renewables,
energy storage, deferrable (schedulable) demand, and demand curtailment (price-responsive demand) Network power flow constraints
Jenkins & Sepulveda (2017) - http://bit.ly/GenXModel
21
17cv01906 Sierra Club v. EPA
ED_001523_00006099-00021
WRESTLING WITH DIMENSIONALIT
TIME
Single Node &
DC Power ` AC Power
Flow
Flow
0
&
0 A
Economic dispatch, no inter-temporal constraints Economic dispatch, ramp & storage constraints
NETWORK
Unit commitment and reserves, binary decisions
OPERATIONAL DETAIL
17cv01906 Sierra Club v. EPA
ED_001523_00006099-00022
Multi-zonal structure with multiple distribution voltage zones and voltage levels
17cv01906 Sierra Club v. EPA
ED_001523_00006099-00023
Siting of resources possible at multiple locations and multiple scales (unit costs)
17cv01906 Sierra Club v. EPA
ED_001523_00006099-00024
Multizonal "pipeline" flow model w/constraints between zones & network expansion
17cv01906 Sierra Club v. EPA
25
ED_001523_00006099-00025
Transmission losses are segment-wise interpolation of losses as a quadratic function of power flows. (Based on Zhang, Hu, & Song 2013 and Fitiwi et al., 2016)
17cv01906 Sierra Club v. EPA
26
ED_001523_00006099-00026
Distribution losses as a function of injections and withdrawals within each voltage level: Semi-urban European network results
Variation in low voltage
Variation in medium voltage
17cv01906 Sierra Club v. EPA
(Jenkins, Luke, & Vergara forthcoming)
ED_001523_00006099-00027
Distribution losses as a function of injections and withdrawals within each voltage level: Semi-urban European network results
Variation in low voltage
Variation in medium voltage
b-- rx
G7 J u t nifi 7,'dHmnnkr +
17cv01906 Sierra Club v. EPA
ED_001523_00006099-00028
Distribution of load curtailment/self-generation necessary to accommodate load growth without network reinforcement - European urban network case
. adi
\ a ? I > d 1 ja f, t d >>v, d
Aiu:
L. i.! t : -aa :h
I' I.f ..<1 i >i I>,*. d
L<-\d ofCurtailim'in ( H'iu-nUion : MW j . ... .
*
r
I I I [ <1 I A
Source: Jenkins, Luke & Vargara, forthcoming
"
17cv01906 Sierra Club v. EPA
29
ED_001523_00006099-00029
Effective low voltage network margin gained (% of initial aggregate peak demand)
Potential for DERs to substitute for distribution network upgrades in representative European distribution networks - low voltage distribution example
10.0%
Semi-urban Urban "^Semi-urban fit
Urban fit
9.0%
8.0%
7.0%
6.0%
5.0%
4.0%
3.0%
2.0%
1.0%
0.0% .....----------------------------------------------- r_-----------------------,-----------------------------r---------------------------- ,
0.00%
0.50%
1.00%
1.50%
2.00%
2.50%
Minimum reduction in aggregate low voltage peak net withdrawal (% of initial aggregate peak demand)
See: Jenkins, Luke & Vergara forthcoming and Prez-Arriaga et al. 2016. The Utility of the Future: an MIT Energy Initiative response to an industry in transition. Cambridge, MA: Massachusetts Institute of Technology. http:/nergy.mit.edu/research/utility-future-study/
3.00%
17cv01906 Sierra Club v. EPA
ED_001523_00006099-00030
GEN-X: a new electricity resource capacity expansion planning model that captures key tradeoffs between locational value and economies of unit scale
17cv01906 Sierra Club v. EPA
ED_001523_00006099-00031
GEN-X APPLICATIONS
5 5
Evaluate economically optimal portfolio of conventional and distributed energy resources to meet projected load growth and policy constraints, as well as implications for network reinforcements
Analyze economics of distributed energy resources and understand how novel resources deliver value and affect operation of power systems
Evaluate optimal scale of solar PV, energy storage devices, fuel cells, etc. (e.g. economies of scale vs. locational benefits tradeoffs)
Evaluating and benchmarking policy and regulation
5
Electricity utility integrated resource planning with distributed energy resources
32
17cv01906 Sierra Club v. EPA
ED_001523_00006099-00032
Jesse D. Jenkins PhD candidate, Institute for Data, Systems & Society Research assistant, MIT Energy Initiative Electric Power Systems Center Massachusetts Institute of Technology jessedj@mit.edu | Linkedn.com/in/jessedjenkins
Questions
17cv01906 Sierra Club v. EPA
ED_001523_00006099-00033
17cv01906 Sierra Club v. EPA
ED_001523_00006099-00034
Megawatts
All capacity - 2035 Spain-like test system, mid-range DER cost declines, transmission constraint case
200,000
150,000
100,000
50,000
0
Transmission expansion annuitized cost ($/MW-yr)
Non-served energy
Lijon - 25MW - 4hr - new
Li_ion -10OkW - 4hr - new
Lijon - 5kW - 4hr - new
Lijon - 25MW - 2hr - new
Lijon -10OkW - 2hr - new
Lijon - 5kW - 2hr - new
Gas turbine - new
Combined cycle gas - new
Solar -100MW - new
Gas turbine - existing
Combined cycle gas - existing
Coal - existing
Nuclear - existing
Solar - existing
^Wind - existing
^--Network expansion
'
17cv01906 Sierra Club v. EPA
ED_001523_00006099-00035
Megawatts
New capacity only - 2035 Spain-like test system, mid-range DER cost declines, transmission constraint case
Non-served energy Li_ion - 25MW - 4hr - new Li_ion -10OkW - 4hr - new Li_ion - 5kW - 4hr - new Li_ion - 25MW - 2hr - new Lijen -10OkW - 2hr - new Li_ion - 5kW - 2hr - new Gas turbine - new Combined cycle gas - new Solar -100MW - new ^--Network expansin
17cv01906 Sierra Club v. EPA
ED_001523_00006099-00036
Megawatts
New capacity only - 2035 Spain-like test system,
mid-range DER cost declines, transmission constraint case
100,000 "i
90,000 80,000
llllilllllliilllllllllj
70,000 60,000 50,000 40,000 30,000 20,000 10,000
I
Effect of losses on locational energy value is insufficient to spur distributed storage alone. If transmission expansion is free, all storage is 25 MW scale at bulk power system level.
Non-served energy Li_ion - 25MW - 4hr - new Li_ion -10OkW - 4hr - new Lijon - 5kW - 4hr - new Li_ion - 25MW - 2hr - new Lijon -10OkW - 2hr - new n Lijon - 5kW - 2hr - new Gas turbine - new Combined cycle gas - new Solar -100MW - new ^--Network expansin
Transmission expansion annuitized cost ($/MW-yr)
17cv01906 Sierra Club v. EPA
ED_001523_00006099-00037
Megawatts
New capacity only - 2035 Spain-like test system,
mid-range DER cost declines, transmission constraint case
100,000
90,000
80,000
70,000
60,000 50,000 40,000
As cost of relieving transmission constraint increases, storage capacity shifts from 25 MW in bulk system to 100 kW in distribution system.
30,000
20,000
10,000
0
Non-served energy Li_ion - 25MW - 4hr - new Li_ion -10OkW - 4hr - new Lijon - 5kW - 4hr - new Li_ion - 25MW - 2hr - new Lijon -10OkW - 2hr - new Lijon - 5kW - 2hr - new Gas turbine - new Combined cycle gas - new Solar -100MW - new ^--Network expansin
Transmission expansion annuitized cost ($/MW-yr)
17cv01906 Sierra Club v. EPA
ED_001523_00006099-00038
Megawatts
New capacity only - 2035 Spain-like test system,
mid-range DER cost declines, transmission constraint case
100,000
90,000
80,000
70,000
60,000
50,000
40,000
30,000
20,000
10,000
0
<3^ P ^CcPP ^CcPP ^CP
N
IP' 5P' <?P'
6S
Transmission expansion annuitized cost ($/MW-yr)
Non-served energy Li_ion - 25MW - 4hr - new Li_ion -10OkW - 4hr - new Lijon - 5kW - 4hr - new Li_ion - 25MW - 2hr - new Lijon -10OkW - 2hr - new Lijon - 5kW - 2hr - new Gas turbine - new Combined cycle gas - new Solar -100MW - new ^--Network expansin
17cv01906 Sierra Club v. EPA
ED_001523_00006099-00039
Megawatts
New capacity only - 2035 Spain-like test system,
mid-range DER cost declines, transmission constraint case
100,000 i
90,000
80,000
70,000
60,000 50,000 40,000 30,000
)ue to declining locational marginal value, torage is not 100% distributed until 'ansmission expansion cost is sufficiently igh. Model avoids incurring "distributed pportumty costs."
20,000
10,000
o
Non-served energy Li_ion - 25MW - 4hr - new Li_ion -10OkW - 4hr - new Lijon - 5kW - 4hr - new Li_ion - 25MW - 2hr - new Lijon -10OkW - 2hr - new n Lijon - 5kW - 2hr - new Gas turbine - new Combined cycle gas - new Solar -100MW - new ^--Network expansin
Transmission expansion annuitized cost ($/MW-yr)
17cv01906 Sierra Club v. EPA
ED_001523_00006099-00040
Megawatts
New capacity only - 2035 Spain-like test system,
mid-range DER cost declines, transmission constraint case
100,000
90,000
80,000
70,000
60,000 50,000 40,000
5 kW storage never deployed as model seeks largest unit size that can capture locational value from transmission capacity deferral.
30,000
20,000
10,000
0
cP
^CPP cP ^ccPP ^CP
oP' <P' v
ip' 5P' sp'
Transmission expansion annuitized cost ($/MW-yr)
Non-served energy Li ion - 25MW - 4hr - new Li ion -100kW - 4hr - new Li ion - 5kW - 4hr - new Li ion - 25MW - 2hr - new Li ion -100kW - 2hr - new Li ion - 5kW - 2hr - new Gas turbine - new Combined cycle gas - new Solar -100MW - new Network expansion
41
17cv01906 Sierra Club v. EPA
ED_001523_00006099-00041
Megawatts
New capacity only - 2035 Spain-like test system,
mid-range DER cost declines, transmission constraint case
100,000
90,000
80,000
70,000
60,000 50,000 40,000 30,000
Changes in locational value have only nodest impact on total capacity of storage or other resources). Ideal location of esources in least-cost portfolio shift, rather ian the composition of the portfolio itself.
20,000
10,000
0
Non-served energy Li_ion - 25MW - 4hr - new Li_ion -10OkW - 4hr - new Lijon - 5kW - 4hr - new Li_ion - 25MW - 2hr - new Lijon -10OkW - 2hr - new n Lijon - 5kW - 2hr - new Gas turbine - new Combined cycle gas - new Solar -100MW - new ^--Network expansin
Transmission expansion annuitized cost ($/MW-yr)
17cv01906 Sierra Club v. EPA
ED_001523_00006099-00042