Energy, Load-Matching & the Building

Start with the two curves, because their shape explains almost everything about on-site solar value. Solar generation over a day is a smooth bell: nothing at night, rising through the morning, peaking near solar noon when the sun is highest, falling to nothing at dusk. It is also seasonal - longer, stronger days in summer, shorter and weaker in winter - and weather-dependent, dropping under cloud. Crucially, the designer cannot move this curve; it is set by the sun, the site and the orientation of the surfaces (a fixed east or west BIPV facade shifts its own smaller peak toward morning or afternoon, but the fundamental daytime-only shape remains).

Building demand follows people and equipment, and its shape depends entirely on the building type. A typical home has a pronounced twin-peak profile: a burst in the morning as people wake, cook and leave, a dip through the middle of the day when the house is often empty, and a long, high evening peak as people return, cook, light, cool and run appliances - a peak that extends well after sunset, when the solar curve has gone to zero. An office or school, by contrast, runs a broad daytime plateau that overlaps beautifully with the solar bell - which is one reason commercial buildings are often better natural matches for solar than homes. A shop, hospital or data centre each has its own signature.

The gap between these two curves is the heart of the matter. In the classic residential case, the building generates a big surplus at midday when nobody is home, and faces its biggest demand in the evening when the panels are dark. This daily mismatch - generation peaking out of phase with demand - is so characteristic of solar-heavy grids that it has a nickname, the 'duck curve'. There is a seasonal version too: in some climates the building generates most in summer but needs most energy (for heating, or for peak cooling) at a different time of year. Understanding the *shape* of both curves for a specific building is the first act of intelligent load-matching, long before any battery is considered.

Once you see the two curves, the key concept follows: self-consumption, the share of generated electricity the building uses itself at the moment it is produced, rather than exporting to the grid. It matters enormously to the economics, and understanding why is central to designing valuable solar.

The reason is an asymmetry in price. When a building consumes its own solar, it avoids buying that unit from the grid - so each self-consumed kilowatt-hour is worth the full retail price of electricity, the expensive rate you would otherwise pay. When a building exports surplus solar, it is usually compensated at a lower rate - under many net-metering and feed-in arrangements, exported units are credited at less than the retail buy price, sometimes much less, and the specifics vary by utility, tariff and country (in India, by state net-metering policy, which a designer must never assume). The result is a simple, powerful rule of thumb: a self-consumed unit is generally worth more than an exported one. Two systems generating identical annual kWh can deliver very different value depending on how much of that generation is used on site versus dumped to the grid at a discount.

This reframes what 'good solar' means. It is not simply the array that generates the most; it is the system whose generation best *coincides* with the building's demand, so that a high fraction is self-consumed at retail value. An office that uses its solar all day may self-consume most of what it makes. A home whose big loads run in the evening may self-consume only a modest fraction of a midday-peaking array and export the rest cheaply - which is why sizing a home array purely to a big annual number, without thinking about self-consumption, can disappoint. The design levers that raise self-consumption are exactly what the rest of this lesson is about: reduce and reshape demand, orient generation toward demand where possible, and - where it pays - store surplus for later. Net metering still matters and export still has value; the point is that self-consumption usually has *more*.

Here is the discipline that separates serious energy design from solar theatre: the first and best move is almost always to reduce the building's demand, not to add more generation. The cheapest, cleanest kilowatt-hour is the one the building never needs in the first place - and a smaller demand is easier and cheaper to meet with on-site solar, needing a smaller array to cover a larger share of it.

Demand reduction is where the interior designer and architect have their greatest leverage, and much of it has nothing to do with PV at all. A well-insulated, well-shaded, thermally sensible envelope cuts the cooling and heating load - the single biggest energy end-use in most buildings. Good daylighting cuts lighting load in the hours the sun is up (and, elegantly, does so exactly when solar is generating). Efficient systems and appliances - efficient air-conditioning, lighting, pumps and equipment - cut the load across the board. Passive-solar and low-energy design (the subject of much of sustainable architecture) do the heavy lifting before a single module is specified. This is why the honest 'net-zero' path, which the next lesson develops, is always reduce-then-generate: you shrink the demand curve first, then size generation to meet what remains.

Skipping this step is the classic, expensive mistake: a building that has done nothing to cut its consumption, then bolts on a large, costly PV system to chase a big generation number, spends far more than one that first halved its demand and met the rest with a modest array. Worse, an inefficient building often has a demand curve badly matched to solar - big evening cooling loads long after the sun is down - so much of the expensive generation is exported cheaply while the building imports at retail in the evening. Reduce demand first, reshape it toward the daytime where you can, and only then size generation to the remainder. For BIPV specifically, this order also keeps the array - the most expensive part - no larger than it needs to be, which matters given that BIPV costs more per watt than bolted-on panels. The binding load calculations and system sizing, of course, belong to the services engineers; the designer owns the demand-first discipline.

With demand reduced, the remaining craft is load-matching: bringing generation and consumption into phase so more solar is self-consumed. There are three broad levers, in rough order of 'do this first'.

The first is design and orientation. Where there is a choice, generating surfaces can be oriented toward when the building needs power: a west-facing array (or facade) shifts its peak toward the evening demand of a home; an east-facing one toward the morning. This trades some total yield (off-optimal orientation, as Lesson 6.1 showed) for better-timed yield - and better-timed yield can be worth more per kWh because more of it is self-consumed at retail value. For BIPV, where facade orientation is often fixed by the building anyway, the honest move is to recognise which facades will generate when, and match uses behind them accordingly.

The second is shifting the demand to meet the supply - load-shifting and, more broadly, demand response. Running discretionary loads during the solar peak - pre-cooling the building in the afternoon, heating water at midday, charging vehicles or running pool pumps and washing machines while the sun is high - moves consumption under the generation curve, raising self-consumption without any battery. In commercial buildings especially, intelligent controls can do a great deal of this automatically.

The third, and the one that most directly conquers the daily mismatch, is storage - typically a battery that charges from the midday surplus and discharges into the evening peak, letting the building self-consume solar it generated hours earlier. Storage can dramatically raise self-consumption and provide resilience, and it is central to the load-matching story. But it must be treated honestly: batteries add significant cost, have their own efficiency losses, degrade over time, and carry real safety and siting considerations - so whether storage pays depends heavily on tariffs, self-consumption gains and local conditions, and the binding sizing, economics, electrical design and fire-safety of any storage system belong squarely to qualified engineers and the codes, never to a designer's assumption. The designer's job is to understand that load-matching - by design, by demand-shifting, and by storage where it pays - is what turns generated kilowatt-hours into real, self-consumed value.