# What a Kilowatt of Avoided Peak Is Actually Worth > The base demand rate is not the answer. Riders, taxes, the ratchet and seasonality all change the figure, usually upward, and the correct number decides every project after it. Section: Cost Models, ROI and Incentives Author: Nil Masferrer Jiménez Published: 2026-08-20 Reading time: 4 min Every demand project on this site produces the same unit of output: one avoided kilowatt of peak. What that unit is worth is therefore the most reused number in the whole exercise, and it is routinely wrong in the same direction. It is worth deriving carefully once. Every appraisal afterward inherits either the accuracy or the error. ## Layer one: everything billed per kilowatt Start by finding every line on the bill priced in dollars per kilowatt, not just the one labeled demand. **Layer one — the all-in demand rate** From one statement and the tariff. - Base demand charge: $14.50 / kW-month - Transmission cost recovery rider: $2.85 / kW-month - Distribution investment rider: $0.95 / kW-month - (Subtotal of demand-based charges): $18.30 / kW-month - Gross receipts and franchise tax at 4.1% of the subtotal: $0.75 / kW-month - **All-in demand rate: $19.05 / kW-month** *31% above the base rate. Every rider quoted in $/kW belongs here, and percentage taxes apply on top of the sum. Rates illustrative.* Building the rider inventory once is the prerequisite, and it is the same inventory needed for every other analysis at the site: [riders and surcharges](https://thedemandcharge.com/articles/riders-surcharges-explained). ## Layer two: the ratchet multiplier If a ratchet clause applies, a peak is not a one-month event. It sets a floor under billing demand for the remainder of the look-back period, so avoiding it avoids every one of those months as well. **Layer two — pricing an excursion under an 80% ratchet** A site whose normal peak is 900 kW. - Normal monthly peak: 900 kW - Excursion peak: 1,300 kW - (Ratchet floor: 80% of 1,300): 1,040 kW - (Excess billed in each following month: 1,040 − 900): 140 kW - All-in demand rate: $19.05 / kW-month - (Cost in the excursion month: 400 × 19.05): $7,620 - (Cost across the following 11 months: 140 × 19.05 × 11): $29,337 - (Total cost of the excursion): $36,957 - **Effective multiple against the single-month figure: 4.85×** *The multiple is specific to the size of the excursion and the ratchet percentage — it is not a constant. Compute it for your own clause rather than borrowing this one. Figures illustrative.* This is where most business cases lose the largest share of the benefit. It also changes what the measure needs to be: under a ratchet, **reliability is worth more than depth**, because one uncovered excursion re-sets the floor and undoes the year: [ratchet clauses](https://thedemandcharge.com/articles/ratchet-clause-explained). ## Layer three: seasonality An annual average conceals a large variation. Where the tariff has seasonal demand rates, a kilowatt avoided in the summer season is worth more than one avoided in winter. Where the ratchet qualifies only on summer peaks, avoiding a summer kilowatt carries the ratchet multiplier and avoiding a winter one does not. Where the demand charge is measured only inside an on-peak window, a kilowatt avoided outside the window is worth nothing at all. The practical consequence is that a measure has to be valued against **the months in which it actually works**. A thermal storage system operating only in the cooling season earns the summer rate for five months and nothing for seven; averaging across twelve gives a number that describes no measure anyone is proposing. ## Layer four: what else moves Three further benefits belong in the figure when they apply, and each is genuinely additional rather than double-counted. **Coincident determinants.** If the measure also reduces demand during the intervals that set a capacity or transmission obligation, that is a separate saving on a separate determinant with its own annual cycle. It needs its own calculation: [coincident and non-coincident demand](https://thedemandcharge.com/articles/coincident-vs-non-coincident-demand). **Energy, where the measure reduces consumption.** Load shifting does not; efficiency does. Storage slightly increases it. Whichever applies, it belongs on its own line rather than blended into a demand figure. **Threshold effects.** If the reduction moves the site below an eligibility threshold for a more favorable schedule, or across the crossover load factor between two schedules, the tariff change is a benefit of the project: [how to choose a rate schedule](https://thedemandcharge.com/articles/how-to-choose-a-rate-schedule). ## Putting it together **The complete figure for one measure** A 180 kW reduction, summer-only, at a site with an 80% ratchet. - All-in demand rate, summer: $19.05 / kW-month - Months the measure is effective: 5 - (Direct demand saving: 180 × 19.05 × 5): $17,145 - Ratchet months protected by avoiding the summer peak: 7 - (Ratchet saving: 180 × 19.05 × 7): $24,003 - Additional energy consumed by the measure, annually: 18,000 kWh - Off-peak energy rate: 5.4 ¢/kWh - (Energy penalty): −$972 - **Annual value of the measure: $40,176** *The ratchet component is larger than the direct saving. A model using the base demand rate for five months would have produced $13,050 — under a third of the real figure. Figures illustrative.* Under a third. That is the difference between a project that clears the hurdle rate and one that never gets built, and it is entirely a matter of reading the tariff correctly. ## The discipline **Deriving your avoided cost per kilowatt** 1. List every charge on the bill priced in $/kW or $/kVA and total them. 2. Apply percentage taxes to that subtotal. 3. Read the ratchet clause and compute the multiplier for a realistic excursion at your site. 4. Identify seasonal rate variation and which months the measure will actually be effective in. 5. Check whether the demand determinant is windowed, and whether the measure acts inside the window. 6. Price any coincident determinant separately. 7. Write the resulting figure down, with the date and the tariff version it came from, and re-derive it after any rate case. That last step matters more than it sounds. The avoided cost is a function of a tariff, and tariffs change. A figure derived three years ago and reused since is a figure describing a rate structure that may no longer exist — and every project appraised against it inherits the drift. Deriving it correctly is the input to the appraisal rather than the appraisal itself; what happens next is in [building the business case for demand reduction](https://thedemandcharge.com/articles/demand-reduction-business-case). ## Sources - [Berkeley Lab — Energy Markets and Policy](https://emp.lbl.gov/) - [DOE Federal Energy Management Program](https://www.energy.gov/femp/federal-energy-management-program) - [EIA — Electricity sales, revenue and average price](https://www.eia.gov/electricity/sales_revenue_price/) - [National Association of Regulatory Utility Commissioners](https://www.naruc.org/) ## Frequently asked questions ### Why is the base demand rate not the right figure? Because it is only part of what an avoided kilowatt avoids. Demand-based riders and percentage taxes stack on top of it, and where a ratchet applies, avoiding a peak also avoids the floor it would have set on later months. ### Does the value change month to month? Frequently yes. Seasonal rates, seasonal ratchet qualification and time-of-use windows all mean a kilowatt avoided in August can be worth several times one avoided in March. ### Should I use an annual average? Only for a rough screen. A measure that reduces the peak in every month earns the annual figure; one that only works in summer earns the summer figure, and averaging conceals the difference. ### What about coincident charges? They need pricing separately, on their own determinant and their own annual cycle. A kilowatt avoided during a coincident peak interval is a different product from a kilowatt avoided during your own monthly maximum. --- Canonical URL: https://thedemandcharge.com/articles/avoided-cost-per-kw Site index for agents: https://thedemandcharge.com/llms.txt