DER Valuation Mihir Desu Strategen Platform 2 About Strategen 3 - - PowerPoint PPT Presentation

DER Valuation

Mihir Desu

Strategen Platform

2

About Strategen

3

Introduction

4

DER philosophy progression

This philosophy progression provides guidance for how all aspects related to DER need to evolve including planning & operation, interconnection, markets & price signals and valuation.

5

Major drivers of storage and distributed resources

6

High renewable penetration in some markets has led to need for additional integration solutions including energy storage. Renewable Energy Penetration In high-population areas where new generation or transmission are expensive to build, storage is serving as a capacity resource.

Local Capacity Needs

Storage is participating in

• rganized

wholesale markets,

• ften by providing

fast frequency regulation services (e.g. PJM Reg-D). Wholesale Market Opportunities Retail customers are utilizing storage as a solution for load shifting and demand charge management. Retail Bill Management Increases in severe weather events impacting grid uptime leads to customers using storage as backup power, especially in the Northeast. Resiliency

7

Possible services and benefits

Current value stack

▪ Energy + losses ▪ No capacity value due to high penetration ▪ No AS value yet ▪ No hedge value to any party because under current policy the energy portion floats to fuel prices

8

~ 10 cents/kWh for HECO

Needed functions

▪ DER energy compensation needs to meet or fall below the fuel rate ▪ DERs need to provide ancillary services and grid support beyond just hosting capacity expansion ▪ DERs need to provide capacity and flexible capacity services ▪ The grid needs to supply enough low cost energy and power to the DER provider so an oversized off-grid system is not justified

9

Future value stack

These services will reduce fixed costs

10

These services will reduce fixed costs in the future Energy should be locked and cost based

Difference between vertically integrated market and competitive market

▪ VDER for competitive? ▪ VOS or RCP for vertically integrated?

11

Valuation Study: Purpose and Goal

1. Gauge if current compensation is fair 2. Figure out appropriate compensation 3. Determining how DERs can help mitigate system costs and benefit ALL ratepayers You regularly buy product X, salesman comes to you and says “I will give you a 100% substitute product for the same price as I forecast it over 25 years?”

1. What would you say? 2. What if there are extra environmental benefits?

12

▪ Generation and distribution capacity benefits of solar, as a non-dispatchable resource, can be small as peak demand hours are generally later in the day, after peak solar generation hours

▪ This is more significant in regions, such as California, that are experiencing a duck curve and negative pricing in the middle of the day

▪ Storage and/or other DERs could help shift solar generation during those later hours to capture all the generation capacity benefits ▪ Creating price signals to induce this type of behavior is critical

Incentivizing Peak Reduction Technologies

13

VDER tariff sends the proper price signals for efficient and effective grid management during peak hours

Illustration of VDER in ISO-NE

VDER Like Approach for New England ISO

Structure:

▪ VDER components are not fixed and instead depend on current year ISO-NE market dynamics (i.e., annual compensation to DERs will vary by year)

▪ Capacity DRIPE value lags 3 years because capacity cleared in the current year’s forward capacity auction (FCA) is not called for 3 years

▪ Certain components of the VDER rate may not be applicable to DERs that are already participating in ISO-NE markets

15

VDER Components

16

▪ Reflects the avoided cost of energy purchases (and avoided line losses) ▪ ISO-NE real-time Nodal LMP Energy Prices (5 min intervals) ▪ Demand reduction induced price effects (DRIPE) ▪ Value of environmental attributes of the generation ▪ Value of RECs to meet RPS requirements or sell into MA SREC market ▪ Impacts on RGGI allowances needed/arbitraged ▪ Value of avoided capacity costs ▪ ISO-NE net regional clearing price * DER’s prior year coincident peak ▪ Demand reduction induced price effects (DRIPE) ▪ Value of avoided transmission system costs due to demand reduction ▪ ISO-NE transmission regional network system (RNS) charges ▪ ISO-NE reliability and administrative charges

Environmental Value Transmission Value Capacity Value Energy Value Distribution Value Ancillary Service Value

▪ Discussed in next section ▪ Reflects the avoided cost of ancillary service purchases ▪ ISO-NE ancillary market charge * DER’s prior year coincident peak

Environmental Value

17

• Renewable Energy Credits (RECs)
• ACP
• MA SREC market
• Regional Greenhouse Gas Initiative (RGGI)

Environmental Value

Transmission Value

18

▪ ISO-NE Regional Network Load (RNL) Charge ▪ Infrastructure: RNS rate (based on annual infrastructure revenue requirement) * IOU’s monthly coincident peak (12-CP) ▪ Reliability: Total ISO-NE payments to resources / RNL monthly peak * IOU’s monthly CP ▪ Administrative: Tariffed rate (based on annual administrative revenue requirement) * IOU’s monthly CP ▪ DER Transmission Value determined by monthly coincident peak * total RNL rate (sum of infrastructure, reliability, & administrative) ▪ Adjusted for IOU-specific line losses

Transmission

Value

Capacity Value

19

▪ ISO-NE Forward Capacity Market (FCM) Charge ▪ Net Regional Clearing Price (NRCP) = Payments made to Capacity Supply Obligations (CSO) / Sum of Capacity Load Obligations (CLO) ▪ FCM Charge = NRCP * IOU’s previous year’s CP ▪ DER Capacity Value = DER reduction of IOU’s CP * NRCP ▪ Note that this value lags a year because of its dependence on the previous year’s CP ▪ DRIPE Capacity Value ▪ DER Capacity DRIPE Value is determined by the reduction in the FCA’s clearing price due to DERs multiplied by amount capacity called in the FCM ▪ This value is difficult to determine because while the change in the FCA’s clearing price due to DERs can be estimated using the FCA’s supply and demand curves, the amount of capacity called in the FCM will not be determined for 3 years

Capacity Value

Energy Value

20

▪ ISO-NE Energy Locational Marginal Price (LMP)

▪ Real-time (RT) Nodal LMPs (5 min intervals) adjusted by IOU- specific line-losses ▪ ISO-NE Net Commitment Period Compensation Charge ▪ Compensates resources for deviations between day-ahead and real-time prices ▪ DRIPE Energy Value ▪ Due to the fact that LMPs are derived incorporating the demand reductions caused by DERs, LMPs are lower than they would be without the presence of DERs ▪ DER Energy DRIPE Value estimated based on average of DRIPE impacts in 2013 and 2015 reports on Avoided Energy Supply Costs (AESC) in New England Energy Value

Ancillary Service (AS) Value

21

▪ ISO-NE Ancillary Market Charge ▪ Regulation Market ▪ Total hourly cost of resource regional compensation / Region’s RT Load Obligation (RTLO) ▪ Forward Zonal Reserves (only during peak hours) ▪ Total hourly cost of zonal resource compensation / Zonal RTLO ▪ RT Zonal Reserves (all hours) ▪ Total hourly cost of zonal resource compensation / Zonal RTLO ▪ Transitional Demand Response (DR) ▪ Total cost of regional resource compensation / Regional RTLO ▪ DER AS Value is estimated by summing the AS rates above (AS Market Charge) and multiplying by the DER reduction of IOU’s pervious year CP ▪ Note that a DRIPE value for ancillary services also theoretically exists

Ancillary Service Value

▪ Assumptions

▪ System Size = 9.28 kWdc ▪ Line Losses = 6.47% (Unitil)1 ▪ Generation Profile = PV Watts hourly ▪ Weather Data = Concord, NH TMY ▪ REC Value = $37.03/MWh2 ▪ LMP = NH Zone

▪ Notes

▪ Does not include DRIPE values for capacity ▪ Nodal LMP will differ depending on actual DER’s location ▪ RGGI value is not easily quantified ▪ Implied value of generation ($/kWh) ~ 0.19

0.00 50.00 100.00 150.00 200.00 250.00

Monthly Revenue ($)

Revenue from VDER Tariff

Energy Capacity Ancillary Services Transmission REC 22

VDER Tariff: Dispatchable Solar DG (1-Year Snapshot)

Rate Comparison for 100% Dispatchable Solar DG (1-Year Snapshot)

50 100 150 200 250

Monthly Revenue ($)

Original NEM Tariff vs. VDER Tariff

Original NEM VDER 23

Approaches to Distribution Valuation

▪ Through rate cases distribution companies know their average marginal system cost ▪ Locational granular values aren’t typically available ▪ Traditional solutions (eg. Transformers, lines) have a known cost and capabilities. The following details are well established:

▪ Timing

▪ Load forecast and work backward with known lead times

▪ Location

▪ Controllable, install equipment in area required

▪ Amount/Capacity

▪ Size and rating of equipment known

▪ Availability

▪ Generally understood but system planning does utilize redundancy for failures

Distribution Valuation Strategies

25

General value vs. locational value

Central Hudson & LNBA Various methods to value grid have been considered in the past Similar to New York DRV process

Distribution Valuation Strategies - NWA Perspective

▪ Timing

▪ Service is only valuable if it is online in time ▪ How do you ensure DER turns up at the right time?

▪ Location

▪ Service is only valuable if it is in the right location ▪ How do you ensure DER turns up at the right location?

▪ Amount/Capacity

▪ Service is only valuable if it is in the right amount ▪ How do you ensure enough DER turns up?

▪ Availability

▪ Service is only valuable if it is available when required ▪ How do you ensure DER is available when required?

26

▪ Timing

▪ Flag pending constraint based on forecasts

▪ Location

▪ Flag an area based on forecast constraints

▪ Amount/Capacity

▪ Detail capacity informed by forecast constraints

▪ Availability

▪ Discussed on next slide

Distribution Valuation Strategies - NWA

27

This information needs to be conveyed to the market Publicly available maps must inform, stimulate and guide market

Project Screening & Selection

28

Is this the only way to capture distribution value?

Comparison of Approaches

▪ LNBA and Central Hudson depend on forecasted constraints and then deferral of planned infrastructure

• projects. Then work backward and calculate an accurate locational value for capacity ($/kW)

▪ The remainder of New York, lacking granular data uses 150% of their MCOS. This is less accurate and granular but does not require know investments nor is it computationally intensive ▪ Central Hudson uses probabilistic load forecast to determine when a planned infrastructure project will be required to develop costs whereas California, (LNBA) does not use probabilistic load forecast at this stage ▪ An effective approach, as performed in New York, is to use the MCOS approach and substitute more granular data when available (bottom up and top down approach)

29

Probabilistic load forecasts Costs developed by dividing cost by capacity required to defer a planned infrastructure project of known value. ($/kW avoided cost) Linear load forecasts Costs developed by considering marginal cost of service ($/kW) and using a 150% multiplier for known constrained areas Load forecasts not directly applicable* Granular location cost details Requires knowledge of proposed investments Significant volumes of data Computational intensive process High-level, approximate location cost details No knowledge of proposed investments required Low volumes of data Non-computational intensive process

Costs are not developed by working backward from a forecast constraint

*

Food for Thought

▪ How much of distribution system cost is linked to demand? ▪ Should a general distribution value be given to rooftop solar?

▪ If so, are you implying that the marginal benefit of solar exceeds:

▪ Call center and billing costs ▪ Poles ▪ Vegetation management ▪ Policy goals ▪ Distribution automation ▪ Cyber security

▪ What is different about energy efficiency?

30

Thank You!

Mihir Desu Manager Strategen Consulting, LLC mdesu@strategen.com

31

Questions?

6th Energy Storage North America (ESNA) Conference + Expo: November 6-8, Pasadena, CA

Largest grid-connected energy storage conference in North America, covering all applications including EV charging (www.esnaexpo.com)

Upcoming 2018 Events

32

Load Carrying Capacity Factor

33

Location Specific Distribution Avoided Costs1

34

Key Steps in Estimating Location Specific Avoided Costs

Load Carrying Capacity Factor

35

Load Carrying Capacity Factor (LCCF) is a measure that captures the ability

• f a given DER to provide effective locational capacity, when and where it

is needed.

Alignment of DERs with highly concentrated peaking risk

Using LCCF to compare individual resources

36

LCCF is a common metric for evaluating locational capacity value across all possible resources, including traditional distribution investments, a variety of single DERs, or an integrated portfolio of DERs.

Building a least cost DER Portfolio

37

In the right combination, DERs deliver more value as a whole solution than as stand-alone components

▪ What if you have too little DER at high value location? ▪ What if you have too much DER at high value location? ▪ What if the DERs fail to perform at high value location? ▪ How do you contract with DERs? Adder? Just exports? ▪ Do utilities need to be prepared to own DER assets?

Policy Considerations

38

It gets complicated fast

Locational System Relief Value

39

▪ Reduction or deferral of costs associated with expanding/replacing/upgrading distribution capacity such as lines, transformers, etc. with the adoption of distributed NEM generation ▪ Costs based on existing estimates for marginal distribution capacity costs as provided by each utility in their Marginal Cost of Service Studies

▪ Adjusted by the expected distribution system peak load reduction value realized by each type of NEM technology based on utility sample substation load data

New York: Distribution Value

40

Method uses distribution marginal COS studies

▪ There are various options to define availability ▪ New York has considered:

▪ Co-incidence between NREL PV production values and Commercial System Relief Program (CSRP)

▪ Also bases incentives on coincidence:

▪ Grid injection during 60 summer hours 2-7pm June-Aug ▪ Grid injection during single highest peak hour ▪ Demand reduction value tied to grid injects during ten peak hours per year

Distribution Valuation Strategies - Availability

41

Distribution Valuation Strategies - NY

▪ DRV

▪ Demand reduction value tied to grid injects during ten peak hours per year

▪ LSRV

▪ Locational adder for high value areas – Same top ten hour target

42

▪ Con Ed - Average system wide cost = $226/kVA ▪ Orange & Rockland - Average system wide cost = $70/kVA ▪ Engineering judgment used to adjust this figure for known constraint areas (150% multiplier, i.e. capacity in constraint areas is estimated to be 50% more valuable than average) ▪ ConEd also estimates typical solar co-incidence DRV and LSRV adjustment

Distribution Valuation Strategies – Distribution Component

43

▪ Central Hudson was the only area to develop location specific values through probabilistic load forecasting methods ▪ Central Hudson Methodology:

▪ Probabilistic load forecasting methodology for granular transmission areas and substations ▪ Analyze load patterns, excess capacity, load growth rates, and the magnitude of expected infrastructure investments at a local level ▪ Develop location specific forecasts of growth with uncertainty ▪ Quantify the probability of any need for infrastructure upgrades at specific locations ▪ Calculate local avoided T&D costs by year and location using probabilistic methods ▪ Identify beneficial locations for DERs

Distribution Valuation Strategies – Central Hudson

44

Distribution Valuation Strategies - NY

45

(MCOS x 150%) – DRV* x 100%

*Central Hudson has specific LSRV values from their study rather than a blanket MCOS x 150% multiplier. Central Hudson’s MCOS=DRV. The remaining area’s DRV are adjusted down from their MSOC to compensate for the higher LSRV areas with a value of 150% MCOS, so the system wide MCOS remains the same.

Locational Net Benefit Analysis

46

▪ Local Net Benefits Analysis (LNBA) ▪ Builds on the DERAC model, DERAC model lacked granular distribution data ▪ Many values are fixed over the year but distribution and transmission deferral and capacity are only valuable during summer where they can defer infrastructure

Distribution Valuation Strategies - CA

47

Distribution Valuation Strategies - LNBA

48

▪ LNBA Distribution capacity takes a very granular view ▪ All specific planned projects in an area, that may be deferred are considered ▪ These are summed to provide a distribution value

Distribution Valuation Strategies - LNBA

49