Cyprus Commercial Solar: Interval Loads Before Array Size

A commercial solar feasibility study should match generation and demand over time before selecting an array size. Annual electricity consumption describes the scale of a Cyprus business, but it does not show whether the site uses electricity during the intervals in which the proposed system produces it. For a 2026 investment decision, interval data, the metering boundary and the applicable connection route should precede the financial model.

This guide presents an auditable method for organising those inputs. Its numerical examples are hypothetical calculations, not measurements of a Cyprus business, tariff quotations or production guarantees. The project engineer must assess the actual electrical design, connection conditions and installation constraints.

Identify the boundary represented by the meter

The study should begin with a drawing or description of the supply being analysed. Identify which buildings, processes and equipment sit behind the meter, and which supplies are separate. A proposed array on a warehouse roof cannot be matched to another building’s demand without checking the physical and contractual arrangement.

Record whether the available data represents import, total site demand or another quantity. On a site with existing generation, measured import alone may not equal the load that equipment actually consumes. The analyst needs to explain any reconstruction of demand and retain the supporting records.

The operating team should identify planned changes before historical data becomes a forecast. An additional production line, altered opening hours or a proposed closure can change the load shape. Label those changes as scenarios with their own assumptions rather than silently editing the historical profile.

Obtain an interpretable interval export

Request the data with its units, timestamp convention and interval definition. A column named consumption may contain energy per interval, average power or a cumulative register. Those formats require different calculations. The export should include enough metadata to identify what each record means.

Check whether timestamps identify the beginning or end of the interval. Confirm the time zone and treatment of daylight-saving changes. A local-time file and a UTC file should be aligned using a consistent conversion, rather than by adding a fixed offset throughout the dataset.

Inspect missing records, duplicated timestamps, estimated values and unusual periods. The method for handling each should be documented. Filling a missing interval with an average may be a practical modelling choice, but it should remain visible as an estimate and be tested where it could affect the decision.

Convert power and energy without confusing them

In a hypothetical 15-minute interval, average site demand of 80 kilowatts corresponds to 20 kilowatt-hours of energy: 80 multiplied by 0.25 hours. Average solar output of 40 kilowatts over that same interval corresponds to 10 kilowatt-hours.

If there is no storage, no other generator and no relevant export constraint, this simplified balance allocates 10 kilowatt-hours of solar energy to direct use and leaves 10 kilowatt-hours of demand to be met from the grid. These values demonstrate the calculation, not a forecast for a specified system.

A different hypothetical interval has average demand of 10 kilowatts and the same average solar output of 40 kilowatts. Demand is then 2.5 kilowatt-hours while generation remains 10 kilowatt-hours. Direct use is 2.5 kilowatt-hours, leaving 7.5 kilowatt-hours for export in the simplified balance. Direct use is 25 per cent of generation for that interval.

The two examples show why an annual total cannot establish direct use. They also assume the stated average powers apply to the same interval. Multiplying an isolated instantaneous power reading by an interval duration without establishing that assumption can misrepresent the energy.

Treat PVGIS as a model with recorded inputs

The European Commission’s Joint Research Centre provides an hourly PVGIS tool that calculates photovoltaic output alongside solar radiation data. Its documentation identifies time in UTC and photovoltaic power output in watts. The output includes configuration information such as location, orientation, nominal power and assumed system losses.

Retain those inputs with the downloaded file. The calculation should be reproducible from the model version, weather dataset and configuration used. A report that gives an annual production total without the underlying assumptions is difficult to assess or update.

Modelled hourly output is not a measured production record from the proposed roof. Local shading, the installed equipment and design constraints require project-specific assessment. If the model’s assumptions do not represent the proposed installation, identify the limitation and obtain a suitable calculation rather than presenting the result as verified output.

Reconcile datasets at a defensible resolution

An hourly generation model and a shorter-interval demand file do not resolve variability in the same way. The analyst should state the common resolution used and how aggregation or interpolation was performed. A finer spreadsheet interval does not create finer information in the source generation model.

For a baseline without storage, allocate direct use as the smaller of demand energy and available generation energy in the aligned interval. Derive the remaining import and export from that balance. Document any constraints that change the available generation or treatment of export.

Review the aggregation effect where short variations matter. An hourly balance can conceal demand and generation that occurred at different moments within that hour. The study should describe that limitation and consider whether additional measurements or a more suitable model are needed for the decision.

Separate operating patterns before averaging

Analyse normal working periods, closed periods and unusual operating conditions separately. A business that runs machinery during daylight hours may have a different solar match from one whose main load occurs outside those hours. The actual interval file should establish that pattern; the sector label alone should not.

Use the operating calendar to explain unusual values. Maintenance shutdowns, equipment faults and temporary opening arrangements can produce a year that is not representative of future operations. Preserve the original data and identify adjustments in a separate scenario.

Present direct use, import and export as separate outputs for each scenario. This allows the commercial team to see whether an array size relies on demand that may change. Avoid describing surplus generation as wasted or valuable until the permitted operating arrangement and applicable terms have been verified.

Check the route before assigning financial values

EAC’s commercial and industrial information describes net-billing installations up to 8 megawatts within its stated framework. That published category is context, not confirmation that a particular site or array is accepted. The applicant should obtain the current procedure and project-specific conditions from the relevant authority.

If the asset will be purchased by a Cyprus company, Sumly's capital-allowances guide is further reading for the equipment-record discussion with its accountant. Keep tax deductions separate from the interval-energy model and obtain an assessment of the actual asset and expenditure.

The financial model should then cite the actual terms used for imports, exports and charges. This guide does not supply a 2026 tariff, grant or export payment. Historical bill values and a proposed contractual arrangement should be labelled with their dates and scopes.

Separate physical results from financial assumptions. A sensitivity analysis can vary verified or clearly hypothetical values without changing the energy calculation. The report should show which assumption drives a result and avoid presenting a single payback figure as a guaranteed outcome.

Evaluate storage as a separate scenario

Storage adds constraints that do not appear in the direct-use examples. The model needs the proposed usable capacity, power limits, efficiency assumptions and operating strategy. It should also account for the permitted connection arrangement and any restriction relevant to charging or export.

A battery cannot be assessed by transferring all surplus energy into a later demand interval without testing those limits. The sequence of charging and discharging matters. Record the state carried between intervals and explain how the analysis treats periods when the battery is full or unavailable.

Compare the storage case against the same baseline and dataset. This makes it possible to identify what the battery changes rather than attributing the entire project benefit to storage. Equipment suitability, safety and backup requirements remain engineering questions outside the arithmetic illustration.

Common questions

Is a year of bills enough to select an array?

Bills help establish billed quantities and charges. They do not necessarily reveal the timing needed to estimate direct use. Obtain suitable interval information and document any remaining limitation.

Can hourly PVGIS values be treated as meter readings?

They are model outputs. Retain their units and configuration and distinguish them from measured site demand or measured photovoltaic production.

What should accompany the feasibility report?

Include the metering boundary, source-data metadata, operating assumptions, model inputs, reconciliation method and verified contractual sources. The report should make clear which results are measured, modelled or hypothetical.