Why Accurate Solar Savings Estimates Need Built-In Utility Rate Databases

A Savings Number Is Only as Good as Its Rate Data

A homeowner almost never asks how many kilowatt-hours an array will make. They ask what it does to the bill. That single question is the entire sale, and it hides a quiet assumption that sinks a lot of proposals: that you can turn production into dollars by multiplying kilowatt-hours by a price. You cannot, because the utility does not bill that way. The bill is the output of a rate structure, and the savings are the difference between the bill the customer pays today and the bill they would pay with solar on the roof.

That means a savings estimate is really two bill calculations run against the same rate schedule, once with the current load and once with the load net of solar production. If the rate schedule in the tool does not match the one printed on the customer's bill, both calculations are wrong, and the difference between two wrong numbers is not a number anyone should sign. The U.S. Energy Information Administration lays out how retail electricity prices reflect generation, transmission, distribution, and regional supply and demand, and utilities package those costs into tariffs that look nothing like a flat per-kilowatt-hour sticker price.

So the accuracy of a savings estimate is downstream of one thing: whether the design tool knows the customer's actual tariff and applies it correctly. Everything else, the layout, the production model, the financing, feeds into a bill calculation that is only as honest as the rate behind it.

The Flat Rate-Per-Kilowatt-Hour Trap

The flat-rate shortcut is tempting because it is easy and because it is close enough to feel responsible. You take the annual bill, divide by annual kilowatt-hours, and get a blended rate of, say, 18 cents. Then you multiply modeled production by that rate and call it savings. It takes thirty seconds and it is wrong in a predictable direction.

The blended rate is an average of a structure, and averages erase the parts that matter. If the customer sits on a tiered rate, their last kilowatt-hours cost far more than their first, and solar shaves off the top tier first, so the true value per avoided kilowatt-hour is higher than the blend early on and then drops. If they are on time-of-use, the blended rate pretends a kilowatt-hour at noon and a kilowatt-hour at 7 p.m. are worth the same, which they are not. And the flat rate assumes every kilowatt-hour the array makes is worth the same whether the home uses it on the spot or sends it to the grid, which under current export rules is almost never true.

The result is a number that reads well in the living room and does not survive the first post-installation bill. On a cash deal that is a trust problem. On a financed deal it is worse, because the lender underwrites to a savings figure, and a figure built on a flat assumption can push the monthly loan payment above the real bill reduction. That gap is where a signed deal falls apart during final review, and it is avoidable.

What a Utility Bill Actually Charges For

Before you can model savings you have to read a bill for what it is, which is a stack of separate charges that each respond to solar in their own way. A residential bill usually carries a fixed customer charge, a volumetric energy charge measured in dollars per kilowatt-hour, and increasingly a set of rules about how exported solar is credited. Many commercial bills add a demand charge billed on the single highest short peak in the month.

The fixed charge does not move when you add solar, so no amount of production touches it. The volumetric charge is where most residential savings come from, but its price per kilowatt-hour may vary by tier, by season, and by time of day. The export rules decide what happens to the kilowatt-hours the home makes but does not use in the moment, and under a net billing tariff those are credited at a wholesale-style rate that can be a fraction of the retail price. The EIA publishes average retail prices by state and sector, which is a useful sanity check on the overall level, but the average cannot tell you how a specific bill is assembled.

Reading the bill this way reframes the estimate. Solar does not reduce a bill by a flat percentage. It reduces specific charges by amounts that depend on when the energy is produced and how the tariff treats it, and a tool that models a bill as one blended rate cannot represent any of that.

How Tiered Rates Move the Estimate

Tiered or inclining-block rates charge more per kilowatt-hour as monthly usage climbs past set thresholds. A household might pay a baseline rate for the first block of usage and a higher rate for everything above it, with a third tier above that in some territories. The intent is conservation, and the effect on solar savings is specific.

Because solar reduces net grid consumption, it peels usage off the top of the stack first. The first kilowatt-hours the array offsets are the most expensive ones the household buys, the top-tier energy, which means early production is worth more than the blended rate suggests. As the system offsets more, it works down into the cheaper lower tiers, and the value of each additional kilowatt-hour falls. A flat-rate model misses both halves of this: it undervalues a small system that only clears the top tier and overvalues a large system whose last kilowatt-hours are displacing baseline-rate energy.

This also changes the sizing conversation. On a steeply tiered rate, a system that just clears the household out of the top tier can capture most of the available savings at a fraction of the cost of full offset, and the proposal should be able to show that trade-off rather than default to maximum coverage. The Department of Energy's homeowner's guide to going solar frames right-sizing around actual consumption and local rules, and tiered pricing is one of the clearest cases where more panels do not mean proportionally more savings.

Why Time-of-Use Reprices Every Kilowatt-Hour

Time-of-use rates attach a different price to the same kilowatt-hour depending on the hour and season. A typical schedule charges a high rate during an evening peak window, a lower rate off-peak overnight, and something in between during the day. Under these rates the question stops being how much the array produces and becomes when it produces relative to when energy is expensive.

This is where a solar estimate gets genuinely subtle. A south-facing array makes the most total energy, but it makes most of it at midday, which on many time-of-use schedules is a mid-priced or off-peak period. A west-facing array makes less total energy and loses a little to the lower winter sun, but it pushes production later into the afternoon and toward the start of the expensive evening window, so each kilowatt-hour it makes can be worth more. The best orientation for total output is not always the best orientation for the bill.

For a homeowner this also interacts with batteries and with the export rules below. Energy produced at midday and stored for use during the evening peak avoids the most expensive retail rate of the day, which can be worth far more than exporting that same midday energy at a low credit. None of this can be modeled with a blended rate, because the blended rate has already averaged the peak and off-peak hours into a single meaningless middle. The EIA notes that retail prices move with the cost of supply and delivery across the day and season, and time-of-use rates are the utility passing that hourly variation straight through to the customer.

Demand Charges and Where They Reach

Demand charges bill the single highest short-interval peak in a billing period, in dollars per kilowatt, independent of total energy used. They have long been a fixture of commercial and industrial bills, and they are starting to appear on some residential and small-business schedules as well. They matter for savings estimates because solar only reduces a demand charge when the array happens to be producing hard at the exact moment the peak occurs.

A facility whose peak lands at two in the afternoon on a clear day can see solar trim that peak. One whose peak lands at seven in the evening, after production has fallen off, will see almost no demand reduction from solar alone, no matter how large the array. A savings model that treats the demand charge as if it falls in proportion to energy offset will overstate savings badly, because it credits a reduction that will not appear on the bill. The honest model shows the demand charge barely moving and points to storage as the tool that can actually discharge into the peak.

This is the clearest example of why the estimate has to carry the real tariff. A demand charge is not a rounding error on a commercial bill. It can be a large share of the total, and getting it wrong in either direction misrepresents the project. A tool that only understands energy charges cannot see this part of the bill at all.

NEM 2.0, NEM 3.0, and Net Billing

The rules for crediting exported solar have changed, and the change is large enough to reshape the entire economics of a residential system. Under older net energy metering, often called NEM 2.0 in California, exported kilowatt-hours were credited at or near the full retail rate, so a kilowatt-hour sent to the grid at noon was worth nearly as much as one consumed on the spot. That symmetry made self-consumption timing almost irrelevant to the savings math.

Net billing, the structure California adopted as its successor tariff and often shorthanded as NEM 3.0, breaks that symmetry. Exported energy is credited at a value closer to what the energy is worth to the grid at that hour, which is frequently a small fraction of the retail rate. The California Public Utilities Commission describes the net billing tariff and its hourly export values in detail. Under this kind of rule, a kilowatt-hour the home uses the instant it is produced is worth the full retail rate, while the same kilowatt-hour exported might be worth a quarter of that or less.

This single change flips several design decisions. Self-consumption becomes the goal, batteries become far more valuable because they let a home use its own midday production during the expensive evening instead of exporting it cheap, and the savings for an identical array can differ sharply depending on which tariff the customer falls under. A tool that still assumes retail-rate export, as many flat and older models quietly do, will overstate savings on a net billing customer by a wide margin. The estimate has to know not just the energy rate but the specific export rule, because that rule now carries much of the value.

Three Ways Tools Handle Rate Data

Design and proposal tools fall into three broad camps in how they handle the tariff, and the camp a tool sits in sets a ceiling on how accurate its savings estimate can be. The table lays out the trade-offs on the dimensions that decide whether the number holds up at closing.

The pattern is consistent: accuracy tracks how faithfully the tool reproduces the real tariff, and the methods that scale across a sales team are the ones where the rate data is maintained once and applied everywhere, not retyped on every deal.

Why a Built-In Rate Database Matters

A built-in utility rate database is simply the tool carrying the actual tariffs, mapped to utilities and schedules, so that when you identify the customer's rate the software can run a real bill calculation rather than a shortcut. The value is not only accuracy on a single proposal. It is accuracy that holds across a whole team and stays correct as rates change.

Consider what manual entry asks of a sales rep. They have to find the current tariff sheet, read it correctly, and type the tiers, seasonal splits, time-of-use windows, and export rules into the tool without error, under the pressure of a live deal, for every proposal. One transposed rate or a missed seasonal change quietly corrupts the savings number, and nobody catches it until the real bill arrives. A central database removes that per-proposal failure point and makes the hard part, keeping up with constant tariff changes, somebody's job once instead of everybody's job every time.

The U.S. Energy Information Administration explains that retail electricity prices are built from the costs of generation, transmission, distribution, and the balance of regional supply and demand, and that these components vary by customer class, by season, and by time of day rather than resolving to one flat price.

U.S. Energy Information Administration, Electricity prices and the factors affecting them

The other half of the value is pairing that tariff with an honest production model. A savings number is a bill calculation, and a bill calculation needs both the rate and the load net of production. The National Renewable Energy Laboratory's PVWatts approach to modeling output from location, orientation, and system losses is the kind of production foundation a savings estimate should rest on, so that the kilowatt-hours going into the bill calculation are themselves defensible before the tariff ever touches them.

How to Evaluate a Tool's Rate Data

When you are judging whether a design or proposal tool can produce a savings estimate you would put your name on, the questions are mostly about the rate data and how it is handled. Work down this list before you trust a number it produces.

• Can the tool look up the customer's specific utility and rate schedule, not just a regional or state average?
• Does it model tiered rates, applying solar against the top tier first the way the bill actually works?
• Does it handle time-of-use periods by hour and season rather than collapsing them into one blended rate?
• Can it represent demand charges for commercial schedules, and does it only credit demand savings when production is high at the peak?
• Does it apply the correct export rule, including net billing where production exported is credited below the retail rate?
• Is the rate database maintained and dated, so you can see when a tariff was last updated?
• Does it run a full bill calculation with and without solar, rather than multiplying production by a single price?
• Can it ingest or model an hourly load profile so timing-sensitive rates are priced against real usage?
• Does the output clearly label savings as a modeled estimate and state the rate and assumptions it used?
• Does the same model carry through to the proposal the customer signs, so the number does not change hands between tools?

Savings-Estimate Mistakes That Surface at Closing

The errors below share a root cause, which is treating the rate as simpler than it is. Each one reads fine in a proposal and then shows up as a gap between the promised savings and the first real bill, which is the worst possible time to discover it.

Multiplying production by a blended rate. This is the headline mistake. It ignores tiers, time-of-use, and export rules at once and almost always overstates savings, because it assumes every kilowatt-hour is worth the household's average price.

Assuming retail-rate export on a net billing customer. Under a net billing tariff, exported energy is credited well below retail. A model that still pays retail for exports can overstate savings sharply on a system sized past daytime self-consumption.

Using a stale tariff. Rates change on their own schedule, and a database or a manually entered rate that is a year out of date models a bill the utility no longer sends. Current rate data is not a nice-to-have.

Ignoring the fixed charge. The customer charge does not move with solar, so a model that treats the whole bill as reducible will overstate the achievable savings by whatever that fixed charge adds up to over the year.

Crediting demand savings that will not appear. On a commercial bill, assuming the demand charge falls with energy offset credits a reduction the array cannot deliver unless it is producing at the peak hour.

Optimizing orientation for output instead of value. Under time-of-use, the orientation that maximizes total kilowatt-hours is not always the one that maximizes bill savings, and a flat-rate model cannot tell the difference.

Letting the number change tools between the estimate and the contract. When the savings figure is calculated in one tool and the proposal is built in another, the two can drift, and the version the customer signs is not the one the model produced.

Modeling Savings Against Real Rate Structures

A savings estimate is the promise the whole sale rests on, so it should be built the way the utility builds a bill: from the real rate structure, applied to real production, calculated as the difference between two bills rather than a single multiplication. That is only practical when the rate data lives inside the design tool and stays current, and when the same model that produces the estimate carries through to the document the customer signs.

This is the case for a built-in utility rate database rather than a convenient blended rate. Enact Design and Proposals models savings against the customer's actual rate structure, so tiered pricing, time-of-use periods, and current export rules are reflected in the number instead of averaged away, and the production side rests on a site-specific model rather than nameplate math. Because the design, the production estimate, and the savings calculation share one model, the proposal a customer sees is the proposal the engineering supports, with assumptions a homeowner or a lender can follow back to the tariff on their own bill. Present every figure as a modeled estimate tied to its stated rate and assumptions, and the estimate stops being a sales number and becomes something that holds up when the first bill lands.

Frequently asked questions

Why do solar savings estimates need a utility rate database?

Because savings are the difference between two bills, and a bill is set by a rate structure, not a flat price per kilowatt-hour. Tiered rates, time-of-use periods, demand charges, and export rules each change what a solar kilowatt-hour is worth, and only a database carrying the customer's actual tariff can run a real bill calculation with and without solar. The EIA explains how retail prices are built from several components that vary by class, season, and time of day, which is exactly the structure a flat rate erases.

What is wrong with using a flat rate per kilowatt-hour for savings?

A flat blended rate averages away the parts of the bill that decide solar's real value, and it usually overstates savings. It ignores that solar offsets expensive top-tier energy first, that time-of-use prices a midday kilowatt-hour differently from an evening one, and that exported energy under net billing is credited below retail. The result is a number that fails against the first real bill. The Department of Energy's homeowner guide frames savings around actual consumption and local rules, not a single average price.

How does NEM 3.0 or net billing change solar savings?

Net billing credits exported energy well below the retail rate, so self-consumption becomes the goal and batteries become far more valuable. Under older net metering, exports were worth nearly the retail rate, so timing barely mattered. The California Public Utilities Commission's net billing tariff values exports closer to their hourly worth to the grid, which can be a fraction of retail. An estimate that still assumes retail-rate export will overstate savings for a net billing customer.

Does a time-of-use rate change the best orientation for panels?

It can. A south-facing array makes the most total energy, but much of it lands at midday, which on many time-of-use schedules is a lower-priced window. A west-facing array makes less total energy while pushing production toward the expensive evening peak, so each kilowatt-hour can be worth more. The EIA notes that prices vary across the day and season, which is why the orientation that maximizes output is not always the one that maximizes savings.

How often do utility rates change, and why does that matter?

Rates change on the utility's own schedule, sometimes more than once a year, so a stale tariff models a bill the utility no longer sends. A manually entered rate or an unmaintained database can drift out of date quietly, and the gap only appears when the customer's real bill arrives. The EIA tracks average retail prices by state and sector over time, which shows how much the overall level moves, and a current built-in rate database is what keeps every estimate aligned with the active tariff.

Can one tool handle both the design and the savings estimate?

Yes, and keeping them in one model is what stops the savings number from drifting between the estimate and the contract. When production, the rate structure, and the bill calculation share a single workflow, the proposal the customer signs is the one the engineering supports. Enact Design and Proposals models savings against the customer's real rate structure and carries that model through to the proposal, with production built on a site-specific estimate rather than nameplate math.

Sources

1. CPUC - Net Billing Tariff (NEM successor)
2. EIA - Electricity prices and factors affecting prices
3. EIA - Electricity sales, revenue, and average price
4. EIA - Average price of electricity (FAQ)
5. DOE - Homeowner's Guide to Going Solar
6. NREL - PVWatts production modeling
7. EnergySage - Solar Panel Cost
8. Enact - Design and proposal platform
9. Enact - Solar software