Business Case for a Flexibility Management System, Part 2: Example Analysis

By John Bonnin

(John Bonnin has 25+ years of industry experience, including 15 years at CPS Energy. He has directed the Day Ahead and Real Time activities of Energy Supply and Market Operations, with a focus on managing market risks to native load customers and optimal dispatch of our power generation fleet. Mr. Bonnin has a Master of Science in Management/Computer Resource Management from Webster University and a Bachelor of Science in Chemistry from Louisiana State University)


Introduction

In Part 1 of this discussion, an overview of the basic elements of a business case for a Flexibility Management System was presented.  It scanned the major cost drivers, the value streams, and other considerations involved in a decision to invest in such a system.  Importantly, the distinction was made that the business case should focus on the incremental value and costs enabled by the system.  Implicit in this is establishing the costs and value of the current operational concept.

In Part 2 of this article, we will look at a hypothetical electrical utility with an active Demand Response (DR) program.  We will analyze the existing DR portfolio and potential new flexibility enabled by the Flexibility Management System.  We will also present estimates of costs for the system as well as the cost for expanding the portfolio.  Finally, we will present the results of a 10-year analysis, plus some sensitivities.

Please note:  this discussion is meant as guidance only.  The data presented, while realistic, is not based on any actual company’s performance.  Similarly, although close to actual, power prices, demand response contract rates, and other costs and figures are not taken from any specific business case.  The capital investment estimates for the Flexibility Management System are only sample data to illustrate the process for developing an investment business case.  Lastly, not all aspects of the costs and benefits were included in this analysis – there are many other charges, value streams, escalation schemes, and other approaches that were omitted for simplicity.  A complete investment business case would delve much more deeply into these areas.

Input Assumptions

Our hypothetical company is called “Townsville Electric Company (TEC),” and is a municipal utility serving the city of Townsville.  It is in a NERC region which has an energy-only market, it has a certificated service territory, and it pays its load-ratio share of grid costs, including transmission cost of service. Here are some key statistics about TEC:

 

  • The Townsville city council has typically granted TEC a 3% rate increase every 5 years
  • TEC uses a standard approach to investment decision analysis. For capital investments, a weighted average cost of capital of 5.75% is used, based on expected interest rates and equity return forecasts.  Costs are usually escalated at GDPIDP (for this example, 5% will be used)
  • TEC has approximately 80,000 residential customers on a thermostat-control program for use by the company to curtail load on high demand days. This program was accomplished as a one-time capital project at a cost of approximately $4 per customer
  • TEC has a number of commercial and industrial customers on various Demand Response (DR) programs, and its most recent incentive payments to these customers for these services averaged $40 per KW-year

Analysis of DR Potential and Expansion Expectations

Currently, TEC has an active DR program for managing peak demand across each summer.  Analyses show this system is realizing substantial cost savings each year (data on this presented further down in this article).  The system is manual, with complex procedures for deployment including some automation, some voice messaging, some automated control signals, and some third-party interaction.

The thermostat program for managing residential peak load in summer is a manual system that is either on or off, with no ability to parse thermostats or pre-cool homes.  The Flexibility Management System is expected to increase the participation from residential customers by 100% to 160,000 customers.  This will, however, necessitate investment in technology to connect these additional 80k customers.

Similarly, TEC has incentive programs for commercial and industrial customers focusing on HVAC and refrigeration controls, control of energy-hungry equipment such as compressors, pumps, and furnaces, and scheduling accommodations for recharging electric vehicles, forklifts, and other rolling stock.  The Flexibility Management System is expected to vastly improve the optimization of this program by offering targeted products and flexible deployment arrangements.  This expected to increase the capacity available from commercial and industrial customers, as shown in the following table:

In summary, although the system as it operates today deploys approximately 31 MW against the peak demand of the company, the Flexibility Management System is expected to increase this to 49 MW.  More importantly, energy curtailment is expected to increase from 82 MWH to 126 MWH per event.

Market Value of Flexibility

There are several streams of value that can be captured.  First, we will focus on grid-related costs that are allocated based on load ratio share.  In markets where this mechanism funds transmission systems investment and maintenance, there is a direct savings to the customers based on reducing coincident peak of the system.

In our TEC example, the grid assesses $57/kw each year to each load serving company, based on its load ratio share, for transmission cost of service (TCOS) and other costs.  History shows 85% effectiveness of deploying complete flexible capacity across the coincident peak due to manual system limitations.  The new Flexibility Management System should improve this effectiveness to 95%.  In addition, an extra 18 MW of DR capacity is expected to be enabled by the system.  The table below summarizes this calculation.

Next, we will look at the energy value of the flexibility.  How many megawatt-hours are curtailed in each demand response event and what is this worth in the wholesale market?  One approach is to analyze the forward curve for on-peak power for the summer seasons when use of this flexibility is most likely.  This is a common approach, but has two drawbacks:

  1. Forward prices for on-peak power usually price out a 16-hour schedule, whereas flexibility deployments are typically 1-4 hours.
  2. Forward curves will not yield the types of extreme and volatile prices sometimes seen in the spot or real-time markets. This is by design – a forward product carries a premium price designed to protect a buyer of exposure to price spikes in the spot market.  In exchange for fixed, reasonable prices across the peak hours, the buyer purchases energy during non-peak hours at a higher price than the spot market.  But as a seller of power (deployment of flexible load assets enables sales of excess energy into the wholesale market) it is desirable to know the value of this energy in the spot market, especially since this is the market into which the energy is likely to be sold.

In this analysis, a combination of forward prices and historical prices will be used to determine this target price.  In any investment business case, analysts should strive to find a value that is acceptable to the analysis team as well as management, since this price is a major driver of the value of the DR capability.  The price settled on should give fair value to the energy because of the time of day it is typically deployed, peak periods in summer.  It should not be over-valued either, as wholesale power markets tend to swing widely year-to-year and even month-to-month.  Here is a table showing one approach:

On a yearly basis, the value of the energy portion of the flexibility portfolio can be estimated.

There is also value to the capacity that the demand response portfolio represents.  In markets that operate a capacity market, the value of the portfolio can be more or less directly measured.  In an energy only market, however, the nearest data point we have is the value placed on energy call options.   The pricing of energy options is driven by several factors, including notice time, duration of delivery, strike price, and delivery point.  Instead of performing a complex options valuation analysis, this discussion will simply use a sample option pricing offer of the type one may receive from an energy trading shop.

Further, a demand response product’s short lead time enhances its value beyond typical option products.  Most options require a day-ahead call, but demand response is basically a real-time call option.  Energy options traded in the marketplace usually have a minimum duration of 8 or 16 hours, whereas TEC has the luxury of setting its duration to 1-3 hours when needed.  An estimate of this increased value is captured in the following table.

Strike price has a role to play in the decision to call on an energy call option.  Organizations must decide whether to utilize the product either to supplement its existing power supply or to mitigate against risk of exposure to spot prices. The higher the strike price, generally the lower the capacity premium.  For demand response products, strike price is not as important, as the decision to deploy flexibility resources can be made on the basis of several factors, such as whether a coincident peak is expected, how well are load forecasts predicting actual load, etc.  Moreover, a comprehensive model for determining the incremental cost of deploying each type of flexibility product would be necessary.  For purposes of this analysis, the interaction of strike price and option premium will not be dwelt upon.

How is the option value expected to increase with the new Flexibility Management System?  Using the estimates from the previous charts, we can evaluate this delta as well.

Summary of Value Streams

We now have a basis for calculating expected future value streams from investing in a Flexibility Management System, to include savings in grid costs, the value of energy, and the option value of the DR capacity.  The 10-year forecast of TEC’s current operations is presented, along with the estimated incremental value enabled by the Flexibility Management System project.

Cost Estimates

First, let us estimate capital costs for the new Flexibility Management System.  These costs can vary widely based on vendor, type of solution (on-premises vs. off-premises), functionality, number of integration points, types of customer programs, size of system, and many other factors.  For this case, an off-premises system is chosen, with no upfront capital cost for the system proper, only annual license costs.  There are still up-front costs, however, in this case are considered capital expenditures.  Included here are some very simple categories and round figure sample costs. (Note the capital cost of the one-time investment to account for doubling the number of residential customers enrolled in the thermostat program.)

The Flexibility Management System will have recurring annual costs.  License fees have been mentioned, but it can be expected that ongoing software customization will be needed, as well as training and administration.

As mentioned above, TEC’s current operation incurs annual costs in two areas:  costs associated with payments and incentives for commercial and industrial customers (program costs), and lost revenue due to curtailment of these customers during periods of high demand.  The business case should capture increases in both these costs enabled by the new system, in addition to both capital and ongoing costs of the new Flexibility Management System itself.

The additional program costs are presented in the table below.

Costs of lost revenue are straightforward to forecast, based on expected energy curtailments and rate structure.

Below is presented a consolidated view of the 10-year forecasts for current and expected costs.  In the expected costs area, the ongoing costs of the new Flexibility Management System are included as well as annual recovery of the capital investment, based on the 10-year time horizon and using TEC’s cost of capital.  The forecast of the systems costs, operations costs, lost revenue, and incremental improvement enabled by the Flexibility Management System are shown, escalating into the future.  Note the increase in Lost Revenue after year 5 as an expected rate increase enters the forecast.

Consolidation of Costs and Value

Now it’s time to show the total picture of costs and benefits as described by this approach.

Several observations are made based on this consolidated look:

  1. This comparison is not a cash flow model, as there are at least two major non-cash aspects to the comparison. The value of capacity is the largest driver of value both now and in the incremental outlook.  The second highest driver of value is savings due to coincident peak reduction.  Neither of these value drivers result in cash flow – Grid Cost savings are simply charges that are never recovered from customers.  Capacity value becomes cash flow only if sold into the market.  The assumption here is the capacity is used to manage risk of exposure to high spot prices; a kind of insurance.  Nevertheless, these two factors represent real value, and should be included explicitly in the analysis.
  2. Non-system related costs are more driven by contracts and incentives to commercial and industrial customers than by the revenue losses.
  3. The current operation is quite effective in capturing value beyond its costs.
  4. The new system looks as if it will deliver about a 50% increase in net value, even after factoring in new system costs, capital recovery, and increased program costs and lost revenue.
  5. The net present value of the investment appears very positive, at least in this simplified model.
  6. The NPV of incremental value captured on a per MW basis can also be evaluated. In this analysis, this figure is just under $1M per MW over the ten-year analysis horizon.   The actual value derived in a “real” business case could be compared with other applicable business metrics to assist decision-making.

Sensitivities

We have performed a high-level screen of TEC’s business case and found positive net present value.  We’ve made reasonable estimates of the costs of the new system, the increased costs of additional participation in flexibility programs to be offered by TEC, and we’ve used market-based data to estimate the incremental value the system will enable.

One of the most useful aspects of building a business case such as this is the ability to model outcomes based on changes in input assumptions.  Now it’s time to analyze two factors that drive the outcome of this screen.

Sensitivity 1 decreased the value expected from both the current system and the forecast returns with the Flexibility Management System.  Both the $/MWH value of energy and capacity are decreased by 50%.  Sensitivity 2 increased the capital cost of the software installation and annual costs of the system by 50%.  The graphic below shows the effect on NPV of these sensitivities.

From this high-level comparison it can be seen that the value drivers of energy and capacity have a pronounced effect on the outcome of the analysis.  Also, increasing the expected costs of the new system do not have as massive an effect on the outcome of the analysis.

Is TEC finished with its analysis at this point?  Not remotely.  Leadership will have to judge whether the investment, in spite of its promise, is in alignment with TEC’s strategic direction.  Leadership must also judge whether the assumptions of higher participations in its programs are reasonable, whether it is viable to assume continued incentive payments to commercial and industrial customers at this level.  TEC’s regulator (presumably the City Council) will have a view on the future of the organization and its strategic investments.  Finally, does the organization have the personnel on staff to drive the changes necessary to both implement and optimize such a system in its day-to-day operation.

Conclusion

It has been shown that developing a business case for investing in a Flexibility Management System requires extensive research, detailed cost estimates, and processes for producing forecasts.  For investment decisions of this kind, it is essential that management has confidence in the input parameters, the forecasts of costs and returns, and the process for assembling a comprehensive picture.

This article was intended to point out a few of the important aspects of this kind of analysis, and is not a complete primer.  Each organization has its own cost and value drivers.  Electricity markets are different in various parts of the world.   The value of both energy and capacity vary widely.  The important focus should be to evaluate all the cost and value drivers that are material to the decision.  As mentioned in Part 1, there are usually many other intangible considerations to factor in as well.  It is hoped that these two articles will serve as a guide to assisting organizations towards this investment decision.