Self-consumption and exports: why two identical homes do not save the same

Almost every saving estimate in circulation does the same sum: it multiplies the estimated annual generation by the tariff you pay today. It is a convenient sum and it produces the most striking figures, but it assumes something that is not true: that every kilowatt-hour your system produces is worth the same.

They are not worth the same. A kilowatt-hour your home uses at the instant it is generated never reaches the grid: you simply do not buy it, and so it saves you the full tariff, with every component that tariff contains. A kilowatt-hour generated when nobody is using it goes out to the grid and becomes a surplus, credited on the bill according to the regulation in force.

That second figure need not equal the first. And there lies the origin of a fairly common disappointment: systems that produce exactly what the design announced, and bills that fall by less than the salesperson promised. Neither the equipment nor the sun failed. What failed was the assumption that producing and saving are the same thing.

The practical consequence is that the saving depends not only on how much energy you generate, but on how much you consume at the same time you generate it. And that is no longer a property of the panels: it is a property of your routine.

The self-consumption rate is the share of generated energy that is used on site, without passing through the grid. If your system produces a hundred units over the day and your home uses forty in those same hours, your self-consumption rate is 40 % and the rest is surplus.

It is the figure that separates a serious saving estimate from a brochure. And yet it is rarely stated. Most proposals present annual generation, the percentage of the bill it would cover and a payback period, without ever saying what share of that generation they assume you will consume directly.

Ask for it in writing. Not as a gesture of distrust, but because it is the only way to compare two proposals: two quotes with equivalent equipment can promise very different savings simply because one assumes an optimistic self-consumption rate and the other does not. If nobody asked you when there are people at home or which equipment you run during the day, the number you were given came from a template.

Take two neighbouring homes, with the same roof, the same monthly consumption and the same system installed. In the first, both adults work away from home and nobody is in between seven in the morning and six in the evening. In the second there is remote work, an older adult at home, and the laundry is done at midday.

The generation of both systems will be practically identical: the two roofs receive the same sun. What changes is where that energy goes. The first house exports most of what it produces and buys back from the grid at night, when the family arrives and the stove, the television and the air conditioning come on. The second consumes an appreciable part of its own generation at the moment it happens.

Solar production has a fairly fixed shape: a bell curve centred on midday. Your consumption, by contrast, has whatever shape your life gives it. The more those two curves overlap, the more each kilowatt-hour you produce is worth.

This disqualifies nobody, and it does not make the first house a poor candidate. It still saves, and with net metering the grid works as its backup. But its saving per kilowatt-hour generated will be lower, and that should be reflected in the estimate from the start, not discovered on the third bill.

The surplus is neither lost nor given away, provided the system is legalised. The bidirectional meter installed by your grid operator records separately what you draw and what you deliver, and the retailer nets both flows when billing the period, applying the rules of Resolución CREG 174 de 2021 for small-scale self-generators.

What is worth understanding is that this netting follows a regulation with its own conditions, and that the specific way surpluses are settled depends on your case and your retailer. It is a legitimate question to ask in writing before signing: exactly how your surpluses will be settled and where that will appear on the bill.

There is one case where they are lost entirely, and it is not unusual: a system already producing whose connection process has not finished. Without a bidirectional meter installed, the energy going out to the grid is recorded by nobody. That is why the project is not finished when the panels are up, but when the new meter is operating.

It is also worth removing an idea that comes up in many early conversations: installing more panels than you need in order to “sell” the surplus. The logic of small-scale self-generation is to cover your own consumption. Extending the system beyond that adds investment today in exchange for energy credited on terms that rarely match the value of the energy you stop buying.

This is the actionable part, and it has an uncommon property: it improves the return on your investment without increasing the investment. Nothing has to be bought; flexible loads have to be moved into daylight hours.

Flexible means you do not care when it happens, as long as it happens. Laundry washed at one in the afternoon comes out the same as laundry washed at nine at night. A pool filtered at midday is filtered the same. A tank pumped at midday serves just as well at night. That is the ground where the gain is made.

Start with what can be automated, not with what requires remembering. A timer on the pump, the washing machine set the night before to start at midday, or a water heater with a configured schedule produce a permanent change; an intention to “remember to use the appliances during the day” lasts three weeks.

And beware of the opposite effect: raising self-consumption by consuming more is not saving. The aim is to shift consumption that already exists, not to create new consumption in order to make use of the sun.

Some consumption cannot be moved. If your house is empty from Monday to Friday and the bulk of the spend is night-time lighting, cooking and air conditioning, no timer solves that.

The intuitive reaction is to think of batteries: store the day's surplus to use at night and take self-consumption close to 100 %. Technically it works. Financially, in a place with a stable grid and net metering, it rarely pays for itself: you are buying an asset with a service life of its own in order to capture the difference between the value of self-consumed and exported energy. Batteries are well justified by backup against outages or by the absence of a grid; as a tool for squeezing out savings, the sums usually do not add up.

The most effective lever in that scenario is a different and less eye-catching one: sizing the system for your real profile rather than for your total consumption. A somewhat smaller system with a higher self-consumption rate can have a better return per peso invested than a large one exporting half of what it produces. It is a conversation worth having with whoever designs your project, because it runs against the natural incentive of anyone selling by installed kilowatt-peak.

All of it comes down to one question you can ask in any sales meeting, and it tidies the conversation up considerably: what share of this generation do you assume I will consume at the moment, and what happens to the saving if that share turns out to be lower?

A solid proposal answers without discomfort, explains the assumption and shows the saving split between what you avoid buying and what is credited. A proposal that values all the generation at the full tariff, having never asked about your routine, is not necessarily lying: it is using a template. But the template does not live in your house.