Time-of-use rates and solar: the timing problem

Two schedules that were never designed to match

A solar array runs on the sun. Output tracks the irradiance falling on the panels, which rises through the morning, peaks near solar noon and falls away through the afternoon. Nothing about that is adjustable: it is where the sun is.

A time-of-use tariff runs on the grid. Peak periods are placed where the system is most stressed and power is most expensive to supply, and on the residential tariffs we read that is the evening, when people come home, cook, run air conditioning into the warmest part of the day, and increasingly charge a car.

The two schedules overlap far less than people assume. By the time the peak period opens, a fixed south-facing array is well past its best. By the time the peak closes, it is producing nothing at all.

That is the whole of the timing problem, and it is why a system that produces exactly the kilowatt-hours its proposal promised can still save less than the proposal implied. The kilowatt-hours arrived at the wrong hours.

What a filed tariff actually says

Rather than describe this in the abstract, we read the filed rate schedules, which are public documents utilities must publish.

The first is a voluntary residential time-of-use schedule. Its title contains the design: peak pricing 4 to 9 p.m. every day. Its own definition of the periods reads: "Summer (service from June 1 through September 30): Peak: 4:00 p.m. to 9:00 p.m. All days. Off-Peak: All other times", and the winter definition is the same window over the remaining months.

The rates on that schedule are $0.52240 per kilowatt-hour at summer peak against $0.39940 off-peak, a spread of about 31%. In winter the spread narrows to roughly 8%.

The second schedule matters far more if you are installing solar in that territory, and it is much sharper. Its peak is "4:00 p.m. to 9:00 p.m. every day including weekends and holidays", with a partial-peak from 3 to 4 p.m. and again from 9 p.m. to midnight. Summer rates are $0.55214 at peak, $0.39026 partial-peak and $0.33358 off-peak. That summer peak is about 1.66 times the off-peak rate.

And here is the sentence that makes this compulsory rather than academic: the schedule states that residential customers billed on the Net Billing Tariff must be served under it, and are not required to have any of the electric technologies otherwise needed to qualify.

Read that plainly. In that territory, a new residential solar customer goes onto the net billing tariff, and the net billing tariff puts them onto the rate schedule with the steepest peak-to-off-peak differential available. The steeper rate is not something you opt into by mistake. It arrives with the solar.

What actually changes the answer

There are three real responses to the timing problem, and they differ enormously in cost and in how much they help.

Shift your consumption. Free, and the most underrated. Running the dishwasher, the laundry, the pool pump and any car charging outside the peak window moves demand into hours your system is covering or, at worst, into cheaper hours. On a schedule where peak is 1.66 times off-peak, moving a load out of the evening is worth about a third of its cost. This is the only response with no capital attached.

Point some panels west. A west-facing array produces less over the year than a south-facing one, because it sees less total sun. But it produces later in the day, pushing output into the front of the evening peak. On a tariff with a wide spread that trade can be worth making, and on a flat rate it never is. Whether it pays is arithmetic about your specific tariff, and it is a design decision that has to be made before installation.

Add storage. The direct answer to the problem: charge from the array during the cheap hours, discharge through the peak. It converts midday generation into evening consumption, which is exactly what the tariff is pricing. It is also the expensive option, and whether it pays depends on the spread, on the battery's cost and on how much of the peak it can actually cover.

What does not help is a larger array. If your problem is that generation and peak pricing do not coincide, adding more generation at the same hours does not fix the mismatch. It produces more of what you already have too much of, at hours you were already covered in.

How common is this?

Nationally, time-varying rates remain a minority arrangement. Our own analysis of federal utility filings puts 11.1% of US residential meters on a time-of-use or other time-varying rate, which is about 15,832,411 households.

That number is smaller than most people expect, and it is not evenly spread: enrolment ranges from most of a state's households to almost none, depending on the state. We publish the state-by-state detail, and the much larger gap between meters that could bill this way and customers actually on such a rate, on a separate page.

The relevant point here is narrower. Whether you are on a time-varying rate today tells you little about whether you will be after you install solar, because in at least one large territory that decision is made for you by the tariff you are moved onto. The question is not "am I on time-of-use", it is "what will I be on once this is interconnected".