Everything about demand charges is more expensive than it first appears, and the ratchet is why.
Without a ratchet, an unusually high fifteen-minute interval costs you one month of demand charge. Annoying, quantifiable, over. With a ratchet, that same interval establishes a floor beneath your billing demand, and you continue to pay against that floor in every subsequent month where your actual demand falls below it — for as long as the look-back window runs.
The interval costs the same. The consequence is multiplied by up to twelve.
The mechanism
A typical clause reads, in substance: billing demand shall be the greater of the maximum demand established in the current billing month, or 80 percent of the highest maximum demand established in the preceding eleven months.
Three variables do the work, and all three vary between tariffs:
- The percentage. Commonly somewhere between 50 and 100 percent of the historical peak.
- The look-back period. Commonly eleven or twelve months, sometimes shorter.
- The qualifying months. Some ratchets consider peaks from any month; others only from defined summer or winter months.
Read your own. Assuming a general rule here is how a business case ends up wrong by a factor.
What it looks like over a year
The shaded portion of each bar is demand you are billed for and did not draw. It exists because of one month, and it persists whether or not anything about the site changes.
The arithmetic
The point is easiest to see by pricing a single excursion.
What one bad interval costs under an 80% ratchet
A site whose normal peak is 900 kW has one 1,300 kW interval in June.
- Normal monthly peak900 kW
- Excursion peak, June1,300 kW
- (Ratchet floor: 80% of 1,300)1,040 kW
- (Billed demand in each later month: greater of 900 and 1,040)1,040 kW
- (Excess billed per later month: 1,040 − 900)140 kW
- All-in demand rate, riders included$19.05 / kW-month
- (Cost of the excursion in June itself: 400 × $19.05)$7,620
- (Cost in each of the following 11 months: 140 × $19.05)$2,667 / month
Total cost of one fifteen-minute interval$36,957
Without a ratchet the same interval would have cost $7,620. The ratchet multiplies it by roughly five. Rates and quantities illustrative — substitute your own.
That multiple is the number missing from most demand reduction business cases, and it usually changes the answer. A measure that looks marginal against a monthly demand charge frequently clears comfortably once the ratchet exposure it prevents is priced properly. The full treatment is in what a kilowatt of avoided peak is actually worth.
What it changes about strategy
A ratchet shifts the objective from reducing the average peak to never having a bad month. Those are different problems with different solutions.
Consistency beats optimization. A control strategy that reduces the typical monthly peak by 10 percent but permits one excursion is worth less than one that reduces it by 5 percent and never permits one. Under a ratchet, the worst month sets the floor for the rest.
Reliability of the measure matters more than its size. A battery that covers 90 percent of events and fails on the tenth allows a floor to be set. That is an argument for conservative sizing and for a fallback: sizing a battery for peak shaving.
One-off events need planning. Commissioning a new line, restarting after a shutdown, a maintenance test that runs everything at once — each is an opportunity to set a floor. Sequencing these is nearly free and is covered in staggered startup.
Demand response participation needs checking. Some programs require a test event demonstrating you can shed load, and a test that ramps everything back up at once afterward is capable of setting the very floor the program was meant to help you avoid. Confirm how the tariff treats demand during and immediately after a called event before enrolling.
Checking your own
- Find the ratchet clause in the rate schedule. Note the percentage, the look-back length, and whether it applies to all months or only defined ones.
- List the last twelve months of measured demand from your bills.
- Compute the floor the clause implies for the current month.
- Compare it with the billing demand actually charged. They should agree.
- Identify which historical month is currently setting the floor, and the month in which that peak rolls out of the window.
- Confirm that peaks which have already left the window are no longer being applied. A floor left standing past its expiry is a straightforward billing error.
- Note whether you are currently paying against the floor or against actual demand. If against the floor, reducing this month's peak saves nothing at all until the floor is gone.
That final line is worth pausing on. If you are billed at the floor, incremental peak reduction has zero immediate value. Cutting from 900 kW to 850 kW while the floor sits at 1,040 changes nothing on the bill. The saving only begins when the setting peak rolls off, and it only persists if no new excursion replaces it.
That is not a reason to stop. It is a reason to time the project so the measure is working and proven before the floor expires, so the first unratcheted month is also the first month of the lower peak. Getting that sequence wrong wastes a year of savings, and it is the kind of thing a monthly demand chart makes obvious — see point twelve of the bill audit checklist.
Where it sits in the tariff
Ratchets live in the demand section of the rate schedule, near the definition of billing demand and immediately after the interval definition. They are also one of the things that differ most between schedules, so a tariff comparison that ignores ratchets is not a comparison at all: how to choose a rate schedule treats it as a first-order term rather than a detail, and the underlying vocabulary is in billing determinants.
