These are written to be read once and used, not to be skimmed. Each one is built on the same public datasets as the calculators — the EIA’s Electric Power Monthly for prices and its Residential Energy Consumption Survey for how households actually use electricity — so the numbers in the prose are the numbers in the tools, computed at build time rather than typed in and left to go stale.
Diagnosis
Working out what is actually happening on your bill, in the order that finds it fastest. These start from the numbers you already have rather than from a list of tips.
- Why is my electric bill so high?
Four causes account for nearly every large electricity bill. Here is how to tell which one is yours in about ten minutes, using the bill you already have.
- Standby power is real, and it is not a quarter of your bill
The arithmetic on vampire power, done properly. What always-on devices actually cost, which ones are worth unplugging, and why the famous statistic does not survive contact with a calculator.
- What actually uses the most electricity in a house
The real ranking, from the EIA’s household survey rather than from intuition. Heating, water and cooling are two thirds of it, and the things people worry about are mostly rounding error.
- Why your winter electric bill triples
Backup resistance heat is the usual answer, and it produces no warning of any kind. How to tell whether it is yours, and what else it could be.
- Why your summer electric bill jumps, and what actually helps
Cooling has the widest regional spread of any household end use — a factor of four across US climate zones. What air conditioning really costs, and which of the standard advice survives contact with the numbers.
- How to cut your electricity bill, ranked by what it is worth
Every saving on this site, computed and put in order. Most published advice is a flat list; this one is sorted by dollars, and the bottom of it is worth almost nothing.
- What it really costs to charge an electric car at home
Cost per mile rather than cost per charge, why Level 1 and Level 2 cost almost exactly the same, and the tariff decision that matters more than either.
Method
How the arithmetic on this site works, why the standard published method is wrong, and how to check any of it yourself.
- Duty cycle: the number that makes every appliance estimate wrong
Why nameplate watts times hours overstates a refrigerator threefold, what a duty cycle actually is, and how to measure your own.
- Watts, amps, volts and kilowatt-hours explained
The four units on every appliance label, what each one actually tells you, and the single conversion that turns any of them into money.
- Time-of-use tariffs: what to move, and what not to bother moving
On a time-of-use rate the same kilowatt-hour can cost three times as much at six in the evening as at midnight. Which loads are worth shifting, and which cannot be.
- How to read your electricity bill
The only two numbers that matter, how to work out what you actually pay per kilowatt-hour, and which lines on the bill you cannot change however carefully you use electricity.
Buying
What measurement equipment can and cannot tell you, and when the honest answer is to spend nothing. These are the pages that carry any commercial links; how the site is funded is stated on each.
- Home energy monitors: what they measure and what they cannot
Whole-home energy monitors put every circuit on a graph. Here is what the three types actually measure, which questions each one can answer, and when a $30 plug meter is the better buy.
- How to measure what one appliance actually costs
A fortnight with a plug-in meter settles any argument about a single appliance. The method, the mistakes that ruin the reading, and how to turn kilowatt-hours into an annual figure.
- Thermostat settings that actually change the bill
Setback, the 3–5% per degree rule, why “leave it steady” is wrong for furnaces and right for heat pumps, and the one setting that silently triples a winter bill.
What is in each one
Why is my electric bill so high?
Every article on this question gives you the same list of twenty tips. None of them tells you which one applies to your house. This one is a diagnostic: it narrows a high bill to a cause using two numbers you already have.
Covers: start with kilowatt-hours, not dollars; the four causes, in the order they are likely; what normal looks like where you live; do the subtraction; what not to bother with first.
Duty cycle: the number that makes every appliance estimate wrong
A refrigerator rated at 150 W does not use 150 W. It uses about 48 W averaged over the day, because the compressor is switched off roughly two thirds of the time. The gap between those two numbers is the reason most published running costs are wrong.
Covers: what the number on the label means; the size of the error; every appliance on this site, both ways; why so many sites get it wrong; measuring your own; what we use, and why.
Standby power is real, and it is not a quarter of your bill
You have read that standby power is 5–10% of a household bill, or a quarter, depending on the article. One watt left drawing costs a fixed and knowable amount per year. Let us just multiply.
Covers: adding up a realistic house; standby, appliance by appliance; where the famous statistic comes from; why standby collapsed, and when; the ones actually worth doing something about; the ones not worth doing anything about.
What actually uses the most electricity in a house
Ask people what uses the most electricity and they say lights, televisions and phone chargers. The survey data says space heating, water heating and air conditioning, and it is not close. Here is the ranking, with the numbers.
Covers: the ranking; why the intuition is wrong; it depends heavily on where you live; how this ranking changed, and why the folklore lags; the single-appliance ranking; what this means you should actually do.
Why your winter electric bill triples
A bill that doubles between November and January is almost always one mechanism, and it is not the one people expect. It is a heat pump quietly falling back on resistance elements that cost three times as much per unit of heat.
Covers: the mechanism; four things that trigger it unnecessarily; how to tell whether it is happening; if you do not have a heat pump; degree days, and why “it was colder” is measurable; fix the settings before the equipment; and check it is not just the rate.
Why your summer electric bill jumps, and what actually helps
Air conditioning is the most geographically variable thing in this entire dataset. The EIA puts it at 4,173 kWh a year in the hot-humid zone against 902 in the marine one. Any single national figure for what cooling costs is describing nowhere.
Covers: what cooling actually costs; the setpoint is most of the lever; which of the standard advice survives; humidity is half the load, and it is invisible; what seer actually buys you; the other summer loads people forget; sizing, and why bigger is not better.
Watts, amps, volts and kilowatt-hours explained
Nearly every mistake in household energy comes from confusing a rate with a quantity. Watts are a rate. Kilowatt-hours are a quantity. Only one of them appears on your bill.
Covers: the four units; the conversions worth memorising; where labels mislead; worked examples; what amps tell you that watts do not; reading your own bill; reading the meter itself; why none of this is on the appliance label.
How to cut your electricity bill, ranked by what it is worth
The problem with energy-saving advice is not that it is wrong. It is that it is presented as a flat list, so unplugging chargers sits beside replacing a water heater as though they were comparable. They differ by a factor of several hundred. Here is the same advice, sorted.
Covers: the ranking; free things first, in order; what is not on this list, and why; when to do each of these.
Time-of-use tariffs: what to move, and what not to bother moving
Every figure on this site uses an average rate, because that is what the published data gives. On a time-of-use tariff there is no single rate — and the difference between the cheapest and dearest hour is often larger than the difference between the cheapest and dearest US state.
Covers: what a time-of-use rate actually is; what is worth shifting; the arithmetic of pre-cooling; should you switch to a time-of-use tariff at all?; a worked example; why this site quotes a flat rate.
How to read your electricity bill
An electricity bill is designed to be paid, not to be understood. But two numbers on it answer nearly every question people have, and one calculation you can do in ten seconds gives you the figure every calculator on this site actually wants.
Covers: the two numbers that matter; working out what you really pay; the lines you cannot change; estimated readings, and the catch-up bill; budget billing, and why it hides things; if you have solar; what to do with the numbers once you have them.
What it really costs to charge an electric car at home
Almost every answer to this question is given per charge, which is useless because batteries are different sizes. The only figure that compares with a tank of petrol is cost per mile, and it is straightforward to compute.
Covers: cost per mile, which is the only comparable number; level 1 and level 2 cost almost the same; the tariff decision is worth more than the charger; home against public charging; do you actually need level 2?; what people forget to count; what the installation costs, and when it is not needed.
Home energy monitors: what they measure and what they cannot
Every figure on this site is a model of a typical appliance. A monitor replaces the model with your house. That is worth real money if you have an unexplained bill, and worth nothing at all if you already know where it goes.
Covers: three kinds, and they answer different questions; what a whole-home monitor genuinely gives you; what they cannot do; when the cheap option is the right one; how to choose; where to look; what to check before you order; what the readings are actually worth.
How to measure what one appliance actually costs
Estimates are useful. A measurement is better, and for a plug-in appliance it costs about the same as a month of running the thing you are measuring.
Covers: the method; the four mistakes that ruin a reading; three worked examples; what a usable meter has to do; the measurement that needs no equipment; what you cannot meter this way; what to buy, if anything; turning the number into a decision.
Thermostat settings that actually change the bill
Heating and cooling are the largest end uses in most homes, so the thermostat is the highest-leverage control in the building. Most of the advice about it is either wrong or right only for one kind of system.
Covers: the one rule that holds everywhere; setback: right for furnaces, complicated for heat pumps; modest setback is the heat-pump answer; cooling: the setpoint beats the equipment; what a setback is actually worth; the balance point, and why it decides your winter bill; what does not work.
If you are here because of a specific bill
Read them in this order. It goes from cheapest to most effortful, and most people find their answer in the first two.
- Why is my electric bill so high — the diagnostic. Start with kilowatt-hours rather than dollars, and rule out a rate change before touching anything in the house.
- Break the bill down — subtract what your end uses should cost from what you actually paid. The remainder is the part worth investigating.
- Standby power — if the remainder is large, this tells you how much of it can plausibly be idle devices, and how much cannot.
- Duty cycle — the concept everything here rests on, and the reason the figures on this site differ from the ones you will find elsewhere.
- Measure one appliance — once you have a suspect, stop estimating and meter it.
Questions about the guides
Do I need to read these to use the calculators?
Are the numbers in the guides the same as the ones in the tools?
Which of these is worth reading if I only read one?
How these are kept honest
Every figure quoted in a guide is computed from the data files at build time — a guide cannot cite a price that the rate table does not also show, because both read the same source. Where a figure is a typical range rather than a measurement, the text says so. The method is written out in full and every dataset is listed with its checksum.