What an Oversight Board Should Ask Before Voting on Rates, Capital, and Large Loads

A board, council, or commission is usually asked to vote on three things that look separate: a rate change, a capital plan, and an agreement to serve a very large customer. They are one decision seen from three sides. Every project a utility builds adds to the revenue requirement, and the revenue requirement is what rates must collect.

This article follows that chain in order and ends with the questions to ask before each vote. The numbers are illustrative figures from our course for non-financial professionals, not from any one utility.

The short version

• Rates are set from cost of service, not from what a market would bear, and only prudent costs belong in them.
• The revenue requirement has four parts: operating expenses, depreciation, taxes, and return on rate base.
• Every capital project that enters rate base adds depreciation and return to rates for the life of the asset.
• A large load agreement is a capital decision. The most consequential term in it is who pays for the infrastructure.

Who decides, and by what standard

One set of poles, wires, and substations serves each area, and customers cannot choose a second set. Regulation stands in for the competition that cannot exist. Rates must be high enough for the utility to stay financially sound, and no higher than necessary to do that.

The standard is the same for every utility. The decision-maker is not:

OwnershipWho decidesHow rates changeInvestor-ownedState commissionFormal rate case, typically 9 to 18 monthsCooperativeMember-elected boardBoard vote after staff analysis; new rates usually 30 to 60 days laterMunicipalCouncil or utility boardOrdinance or resolution, usually after a public hearing

A cooperative or municipal board has no commission reviewing its work. That increases its responsibility. The board or council is the only check on whether rates are adequate and fairly designed.

Step 1: The revenue requirement

The revenue requirement is the total a utility must collect to stay financially sound. For an illustrative cooperative it looks like this:

ComponentAnnual amountOperating expenses$7,300,000Depreciation$1,920,000Taxes (property)$380,000Return on rate base ($44.7 million at 6.0%)$2,682,000Revenue requirement$12,282,000

If current rates produce $11,400,000, the deficiency is $882,000, about 7.7% more than current revenue. Deferring the decision does not close that gap. It grows, and so does the increase needed later.

Four things push the requirement up: operating costs (a 5% increase on $4.4 million of purchased power adds $220,000), new capital, new debt, and declining sales that spread the same total over fewer kWh.

Step 2: Rate base and the capital project

Rate base is the net value of the plant on which the utility earns a return. To belong in it, an asset must be used and useful, and the investment must have been prudent. Plant that customers or developers paid for directly, called contributions in aid of construction (CIAC), stays out, so customers do not pay twice.

Consider a $4.2 million substation at a 6.0% return and a 35-year life:

• Return: $4.2 million x 6.0% = $252,000 a year.
• Depreciation: $4.2 million over 35 years = $120,000 a year.
• Added to rates: about $372,000 in the first year.

The return portion declines as the plant depreciates, but the effect lasts for decades. A customer contribution reduces both amounts. This is why a capital approval is also a rate decision. I covered the large load version of it in Rate Base Implications of Large Load Agreements.

Step 3: Who pays, and in what form

The requirement answers how much. Rate design answers who pays and in what form. A cost of service study assigns each customer class its share, and the design converts that share into fixed, energy, and demand charges. Each charge follows a different cost: the number of customers, the kWh used, and the peak kW the system must be built to carry.

Large industrial bills are often driven by demand, commonly 40 to 70% of the bill. Rate design is a policy choice, and every design trades off cost recovery, fairness among classes, price signals, and stable bills. Someone always subsidizes someone, and what counts as fair is contested.

Step 4: The capital plan

Capital spending is not optional for most utilities. Poles, wires, and substations last 25 to 50 years, and aging plant, reliability obligations, and load growth drive most of the budget. Deferring a project to hold rates flat does not remove the need. In one illustrative substation case, the planned project costs $2.4 million, the same project deferred until it fails costs $4.8 million three years later, and an emergency replacement can reach $7.2 million.

A five-year plan should be read by category and converted into a rate impact. In an illustrative cooperative plan:

• Gross additions total $22.66 million over five years, against an existing rate base of $31.2 million.
• Year 2 is a one-time spike of $8.02 million from a major substation plus solar and batteries.
• At roughly 9% combined depreciation and return per dollar of new rate base, the base plan needs about an 8% rate increase over five years. A reduced plan needs about 5%, and an accelerated plan about 10%.

Approving capital without the matching rate model commits the board to spending without committing it to the revenue that funds the spending. Resource plans work the same way: owned generation enters rate base and creates a permanent increase, while a power purchase agreement is an expense with contract risk instead.

Step 5: Large loads

One large customer can be bigger than the whole existing system. A standard commercial account peaks at 50 to 200 kW. A hyperscale data center can peak at 50 to 500 MW, with a load factor of 85 to 99% against 25 to 40% for the commercial account.

Each large load agreement answers three questions: where and how much, who pays, and what happens if the customer leaves. There are three ways to answer the second one:

1. All ratepayers. The utility builds it and the cost enters rate base. Members subsidize one customer.
2. Hybrid. What benefits the whole system goes in rates, and what serves only this customer is paid by the customer.
3. Customer pays in full (CIAC). The plant stays out of rate base, so members owe nothing on it if the customer leaves.

The exposure is easy to measure. Suppose an 80 MW customer needs $18 million of new plant, of which $14 million serves only that customer. If the customer leaves and the plant is stranded, $14 million at about 9% is $1.26 million a year. Across 8,000 members that is about $157.50 per member per year, for the life of the plant. I wrote about the contract side of this in Stranded Asset Risk in Large Load Agreements.

Six protections for members

1. A binding agreement signed before construction. A letter of intent is not enough.
2. CIAC for dedicated plant, and a deposit or other security before ground is broken.
3. Take-or-pay minimum revenue if the customer under-uses service.
4. Departure terms: the customer repays unrecovered utility-funded plant if it leaves early.
5. Phased construction tied to verified demand milestones.
6. A calculation of the per-member cost if the plant is stranded.

The rate also has to cover what the customer requires, not only what it pays. Many utilities try to serve a large load under an ordinary commercial or industrial tariff. A tariff built for loads up to 5 MW cannot be assumed to work for an 80 MW customer that is 16 times larger. I explain that trap in Why a Standard Large-Power Tariff Isn't Enough.

What FERC has and has not done

In June 2026 FERC issued show cause orders to six regional grid operators on large load interconnection. It did not issue a final rule, and it left retail cost allocation to the states. Your board still decides what your own members pay. The orders and their status are covered in What FERC's June 18 Orders Mean for Utility Rate Design and Interconnection, and the docket should be checked for the latest.

A board does not need a federal rule to ask for the same protections. The customer that triggers an upgrade should bear its cost, and your duty to existing members already requires it.

Questions to ask before you vote

On a rate change

• What is the full revenue requirement, and what is the monthly impact on a home using 900 kWh?
• When was the last cost of service study?
• What do TIER and debt service coverage look like in years 1, 2, and 3, and if purchased power rises 8%?
• Is the increase adequate for 3 to 5 years, or does it invite a second increase next year?

On a capital plan

• What share is non-discretionary, and what does a two-year deferral risk?
• What rate increase does the plan require in each year?
• Was it competitively bid, and is the debt sustainable with a stress case?

On a large load agreement

• What is the total infrastructure, and who funds how much?
• Is there a binding agreement and posted security?
• If the customer leaves in year 3, what do members pay, and for how long?
• Was the rate built from a large load analysis rather than a standard tariff?
• Can we explain this to members at a public meeting?

A body that cannot answer these before voting has not yet met its duty of care. Economic development and member protection both matter, and the answer to a large load request is usually yes, with the right provisions.

The course

I built this material into a course, Rates, Capital Planning, and Large Loads for Non-Financial Professionals. It has 12 short modules written for board members, council members, commissioners, attorneys, and regulatory staff who do not have an accounting background. The course ends with a final exam, and a score of 70% is required to pass. The certificate of completion can be downloaded as soon as you pass.