Real-World Solar Savings vs Projected Savings: How We Build a Credible Forecast
Author: Steve Fairless
Originally Published: 7th June 2026 · Updated: 2nd September 2026
A solar savings projection is a financial model, not a promise about future bills. Real savings can be higher or lower because weather, electricity use, import prices, export tariffs, battery behaviour, downtime and household habits change after the quotation is produced. A credible model makes those assumptions visible and separates the value of electricity used on site from the value of electricity exported.
At Sustainable Energy Engineering, we do not judge a proposal by the largest lifetime-saving headline. We want to know how the annual generation was estimated, when the customer uses electricity, how self-consumption was calculated, what tariff values were assumed and how sensitive the result is to change.
This guide explains the difference between generation and savings, shows how the main calculations work, identifies the assumptions most likely to move the result and sets out what customers should compare when reviewing solar quotes.
Projected vs Real Solar Savings: The Short Answer
- Generation and savings are different. A system can meet its kWh forecast but save less money if more energy is exported than expected.
- Self-consumed solar is usually valued at avoided import cost. Exported solar is valued at the applicable export tariff.
- Weather varies from year to year. MCS requires the standard estimate to be presented as first-year guidance rather than a guarantee.
- Tariff assumptions can dominate long forecasts. Small annual price-growth assumptions compound significantly over 20–25 years.
- Batteries change timing and add losses. Storage may increase self-consumption, but its capacity, power, efficiency, control strategy and replacement cost matter.
- A range is more honest than one perfect number. We prefer conservative, central and favourable scenarios with the assumptions shown.
First Separate Energy from Money
The solar array produces electrical energy measured in kilowatt-hours. The money saved depends on what happens to each kilowatt-hour. Electricity used immediately in the property avoids a grid import. Electricity stored may avoid a later import after battery losses. Electricity exported earns the relevant tariff payment.
Any model should then account for standing charges that remain payable, battery charging from the grid, maintenance, finance costs and expected component replacement where they are relevant to the comparison.
Solar Does Not Normally Remove the Standing Charge
A grid-connected customer usually continues to pay a daily standing charge even when annual solar generation is high. A forecast that treats the entire existing electricity bill as avoidable can overstate the saving.
How Annual Solar Generation Is Estimated
The current MCS solar standard requires a standard estimate based on installed capacity, regional yield, orientation, inclination and shading. It also requires the customer to be told that performance varies with solar radiation from location to location and year to year.
PVGIS provides a second useful perspective because it models monthly and annual production using long-term solar and weather data for a selected location and system configuration. We use models to form a defensible expectation, not to create an exact weather forecast decades into the future.
Generation Inputs We Want to See
- the installed DC capacity in kWp and exact module count;
- orientation and inclination for each roof plane;
- a shade factor or site-specific shading method;
- system losses and inverter assumptions;
- monthly as well as annual generation where possible;
- any export limitation or curtailment expected;
- clear treatment of degradation over the modelled period.
Self-Consumption Is Often the Most Important Financial Assumption
Self-consumption is the share of generated solar used within the property rather than exported. It is shaped by occupancy, appliance use, working patterns, heating, hot water, EV charging and battery operation.
A home empty through the middle of the day may export more than a home with daytime loads. A school or business can use a high share directly if its operating hours align with solar generation. That is why a generic self-consumption percentage should not be applied without explaining the customer profile.
For domestic solar PV, we ask for annual use and, where available, half-hourly smart-meter data. This helps us distinguish a high annual consumer with evening demand from a lower annual consumer whose demand overlaps strongly with daytime generation.
Why Export and Import Need Different Values
The Smart Export Guarantee gives eligible small-scale generators a route to payment, but suppliers set their own rates, contract lengths and terms. Export is therefore not a fixed national value.
One exported kilowatt-hour should be valued at the expected export rate, not the retail import rate. If an installer values every generated unit as though it replaces a full-price import, the forecast can materially overstate the financial benefit.
| Energy route | Financial treatment | Main uncertainty |
|---|---|---|
| Used directly | Avoids buying a similar kWh from the grid at that time. | Actual overlap between solar output and household demand. |
| Stored and used later | Avoids a later import after battery losses and reserve settings. | Battery throughput, efficiency, power limits and control. |
| Exported | Earns the supplier’s export rate for metered export. | Tariff eligibility, future rates and export timing. |
| Curtailed | Produces no direct financial value. | Export limits, full batteries and low on-site demand. |
Electricity-Price Inflation Can Transform a 25-Year Headline
Long-term models often assume that electricity prices rise every year. Compounding makes even a modest percentage produce a much higher future unit rate. That may happen, but no installer knows the exact retail price in year 18.
We prefer to show more than one scenario:
Conservative Scenario
- flat or weak real energy-price growth;
- current export terms reduced or held cautiously;
- lower self-consumption;
- allowance for component replacement.
Favourable Scenario
- higher avoided import value;
- strong load matching;
- useful export or smart-tariff value;
- good system availability.
The central forecast should not depend on the favourable case. A system that remains sensible under cautious assumptions is easier to recommend with confidence.
An Illustrative One-Year Comparison
Consider a system forecast to generate 4,000 kWh in a year. The following examples use invented tariff values to demonstrate the method; they are not a quotation or current national average.
| Scenario | Used on site | Exported | Illustrative annual value |
|---|---|---|---|
| Low self-consumption | 1,200 kWh at 28p = £336 | 2,800 kWh at 12p = £336 | £672 |
| Balanced | 2,000 kWh at 28p = £560 | 2,000 kWh at 12p = £240 | £800 |
| High self-consumption | 3,000 kWh at 28p = £840 | 1,000 kWh at 12p = £120 | £960 |
The panels generated the same amount in all three cases. The £288 gap came from energy timing and valuation. This is why usage assumptions can matter more than a small difference in module efficiency.
How Batteries Change the Forecast
A battery can increase the share of solar used on site by storing midday surplus for later. The model must not assume that every exported kilowatt-hour can be recovered. Battery capacity, charge power, discharge power, minimum reserve, efficiency and seasonal surplus limit the result.
A battery may also charge from a cheap time-of-use tariff. That can create value beyond solar self-consumption, but the calculation should separate solar savings from tariff arbitrage so the customer understands where the return originates.
When we design solar PV with battery storage, we compare the battery with a PV-only baseline. That prevents the value created by the panels from being counted again as though it were produced entirely by the battery.
Battery Replacement and Degradation Belong in a Long Model
Energy Saving Trust notes that a typical battery may last around 10 to 12 years, while solar panels commonly operate for much longer. A 25-year cashflow should therefore state whether it assumes battery replacement and at what cost.
The same principle applies to inverters. A long-term forecast that counts 25 years of savings but no likely equipment replacement, maintenance or downtime is incomplete.
Finance Cost Can Change the Payback
Simple payback divides the upfront cost by annual savings and ignores the cost of borrowing, changes in savings, maintenance and the time value of money. If the system is financed, interest and fees affect the household cashflow.
That formula is useful as a quick indicator, but it should not be confused with net present value or an investment return calculation. We explain which method is being shown so customers can compare like with like.
Weather Variation Is Normal
A below-forecast month does not automatically mean a fault. Solar radiation can vary substantially from one month or year to another. Performance should be compared over a sensible period and with local weather context.
A persistent drop, inverter alarm, failed string or unexplained difference between similar roof planes deserves investigation. The design forecast is a benchmark, but monitoring and diagnosis establish why the actual result moved.
Behaviour Can Improve or Reduce Savings
Customers often change how they use electricity after solar is installed. Running a dishwasher, washing machine, immersion heater or EV during productive hours can increase direct use. Adding a heat pump or working from home changes demand more fundamentally.
Some behaviour changes are beneficial; others simply shift demand into a period when solar is still too low. Monitoring should confirm the result rather than assuming every timed appliance is powered entirely by the roof.
Smart Tariffs Add Opportunity and Complexity
Time-of-use tariffs can make battery operation more valuable, but they also make the model more sensitive to control settings and tariff changes. Import avoided at a peak rate is worth more than import avoided at a cheap overnight rate. Export may also have time-dependent value.
Our energy tariff guidance focuses on the full import/export structure. We do not recommend a battery strategy from one headline rate while ignoring standing charges, eligibility, compatibility and the hours when the customer actually imports.
Common Forecasting Errors We Look For
Warning Signs in a Solar Savings Proposal
- every generated kWh is valued at the full import rate;
- self-consumption is high but no usage profile or battery method is shown;
- export income and avoided import are counted on the same energy;
- electricity-price inflation is aggressive and presented as certain;
- the model omits shading, degradation or export limitation;
- battery losses, reserve and replacement are ignored;
- standing charges are treated as avoidable;
- finance interest is excluded from a financed-system comparison;
- the lifetime total is prominent but the first-year calculation is unclear;
- no sensitivity range is provided.
What We Ask Customers to Provide
Better input data produces a better forecast. We want recent electricity bills, annual kWh use, current tariff details and half-hourly consumption where available. We also ask about future EVs, heat pumps, electric hot water, extensions, home working and changes in occupancy.
The roof survey then adds orientation, pitch, shading, available area, mounting constraints and electrical information. Financial modelling should follow the physical design, not lead it.
How to Compare Two Quotes Fairly
- Compare installed capacity and expected annual generation. Check that shade and roof geometry are treated consistently.
- Compare self-consumption assumptions. Ask what demand profile supports each percentage.
- Separate import saving and export income. Confirm that the same energy is not counted twice.
- Check battery assumptions. Look for usable capacity, losses, power limits, reserve and replacement.
- Review tariff and inflation assumptions. Identify current values and long-term escalation.
- Include total installed and finance cost. Compare cash price, interest and any ongoing fees.
- Look beyond payback. Consider warranties, aftercare, monitoring, serviceability and system life.
Our Conclusion: A Forecast Should Explain the Result, Not Sell the Result
A strong solar projection is transparent enough for the customer to understand why the saving changes when generation, self-consumption, import price or export value changes. It acknowledges uncertainty without making the forecast meaningless.
We prefer realistic assumptions, clear first-year figures and sensitivity ranges. That approach may produce a less dramatic headline than an optimistic sales model, but it gives the customer a much stronger basis for deciding whether the system fits their property and financial priorities.
Want a Solar Forecast Built Around Your Actual Energy Use?
Provide your annual consumption, tariff and any future EV, heat pump or battery plans. We can model the roof, generation, self-consumption and export without hiding the assumptions.
