Industrial photovoltaics – how to match a PV system to the actual energy consumption profile?

Industrial photovoltaics should not start with the question of how many panels will fit on the roof. In a manufacturing plant, it is more important to know when electricity is consumed, when the PV system will generate it and what share of that generation will be used on site. Only then can you assess what system size makes sense, where to locate it and whether it is worth combining it with energy storage.

Photovoltaic modules generate direct current, and the inverter converts it into alternating current used by the plant’s electrical system. The technology itself, however, is only one part of the investment. In industry, the project outcome depends primarily on how well the generation source is matched to the company’s overall energy system.

In brief

• Annual electricity consumption in MWh or GWh does not show whether the plant will use PV electricity during the same hours in which it is generated.
• The analysis should use an hourly load profile, and preferably available 15-minute data, covering a representative year together with the production calendar.
• Working days, weekends, process shutdowns and seasonality can significantly change the level of self-consumption.
• Available roof or ground area defines the technical potential, but it should not by itself determine the size of the investment.
• Connection conditions and the ability to export surplus electricity to the grid should be checked before the concept is finalised.
• Energy storage makes sense when it has a specific function: for example, shifting surplus PV generation to other hours, limiting exports, reducing demand peaks or providing backup power in a properly designed system.

Industrial photovoltaics starts with the plant’s operating profile, not with roof area

Industrial photovoltaic installations are often initially assessed using two figures: annual electricity consumption and the area available for modules. Both are useful, but neither shows what matters most: the coincidence of PV generation and the plant’s electricity demand.

A photovoltaic system generates electricity mainly during the day, and its output profile changes with the season and weather conditions. A plant, meanwhile, may operate one, two or three shifts, stop production at weekends, run seasonal production campaigns or maintain a high baseload around the clock. Two companies with identical annual consumption can therefore have completely different potential to use the same PV system.

Industrial photovoltaics therefore requires time-series analysis rather than just an annual energy balance. Production can be modelled, among other sources, using hourly PVGIS data, which allow photovoltaic electricity generation to be analysed hour by hour. For an industrial plant, this generation profile then needs to be compared with the actual electricity demand profile.

Why can two companies with the same annual consumption need different PV systems?

Assume that two plants each consume 10 GWh of electricity per year. The figure of 10 GWh alone does not tell us which photovoltaic option is appropriate.

At Plant A, PV electricity may cover daytime needs of the production lines well, but on Saturday and Sunday the same system may generate larger surpluses. Plant B has the same total consumption, but a substantial share occurs at night, when photovoltaics does not generate electricity. As a result, the same annual energy balance does not automatically lead to the same system capacity, configuration or economics.

This also shows why simple rules used for small installations should not be transferred directly to a large industrial plant. Industrial analysis must account for the production rhythm, the structure of loads and infrastructure constraints.

What does the hourly load profile show that an annual bill does not?

Invoices are a good starting point for checking electricity consumption and costs, but a monthly total does not show how the load changes over time. In practice, metering data should be collected at the shortest reasonably available interval – 15-minute electricity data are often available – and at least one representative full year should be reviewed.

Such a profile can answer questions including:

• what is the plant’s minimum baseload during the day and at night;
• how does the load change between production shifts;
• do demand peaks occur during PV generation hours;
• what do Saturdays, Sundays and public holidays look like;
• when do maintenance and process shutdowns occur;
• does consumption increase in summer, winter or during specific production campaigns;
• what share of potential PV generation can be consumed directly on site.

If the plant does not have well-organised data, it is worth improving energy monitoring and management first. The larger the installation and the more variable the process, the more important the quality of the input data becomes.

From an auditor’s perspective, I would not start by selecting modules or inverters. I would first examine a typical operating week, how individual months differ and which loads form the daytime baseload. Only then is it worth comparing technical options.

Working days, weekends and process shutdowns

At a plant operating Monday to Friday, the PV profile may match production well in the middle of the week but behave very differently at the weekend. A large system designed only around peak demand on a working day may generate a surplus on Saturday that the plant cannot use.

The same applies to longer shutdowns. If a company has an annual maintenance shutdown in July or August, it should be overlaid on the PV generation profile. A period of very strong solar generation may coincide with the plant’s lowest electricity demand.

At minimum, it is worth analysing separately:

• a typical working day;
• a typical Saturday and Sunday;
• days with reduced production;
• planned shutdowns;
• start-up periods and periods of increased load.

In the office area, demand often aligns with daylight hours better than a process that runs mainly at night. In a large company, however, offices should not be analysed separately from production halls, warehouses, refrigeration, compressed air and other loads operating according to their own schedules.

Production seasonality versus PV seasonality

Photovoltaics has its own seasonality: monthly and hourly generation changes throughout the year. PVGIS provides both monthly and hourly data, so an industrial analysis can compare variations in PV generation with the plant’s operating calendar.

If a company uses the most electricity in summer – for example because of intensive production, cooling or air conditioning – the coincidence with PV generation may be favourable. At another plant, peak consumption may occur in winter while production lines run at lower loads in summer. Once again, the annual energy total does not reveal this difference.

Planned changes are also important. A new production line, process electrification, expanded refrigeration, vehicle chargers or a different shift system may change the profile after the PV system has been commissioned. The analysis should therefore include not only historical data but also a credible plant development plan.

Self-consumption in industrial photovoltaics

Self-consumption is the share of electricity generated by the PV system that is used on site at the same time. In a large plant, it is one of the key investment parameters because electricity consumed directly replaces part of the electricity imported from the grid.

This does not mean that the goal should always be to maximise self-consumption at any cost. The right option depends on the value of electricity consumed on site, the terms for selling or settling surplus generation, investment cost, technical constraints and planned changes at the plant.

The analysis should distinguish between three quantities:

1. electricity generated by the PV system;
2. electricity used directly on site;
3. electricity exported to the grid or sent to storage.

This makes it possible to determine whether the issue is excessive generation at certain hours, insufficient plant demand or a limit on export capacity.

Connection capacity and grid constraints can change the design

In industrial investments, demand-side parameters must be distinguished from generation-side parameters. The site’s connection capacity and contracted demand on the consumption side do not automatically constitute permission to export the same PV capacity to the grid. The way a generating source is connected depends, among other things, on its capacity, configuration and the requirements of the relevant network operator.

Before the concept is finalised, the connection procedure applicable to the specific generating source therefore needs to be established, and the existing transformer station, switchgear, protection systems, metering arrangement and permitted mode of parallel operation with the grid should be reviewed. Requirements depend on project capacity and configuration and on the relevant operator. For example, the current TAURON Dystrybucja forms distinguish documentation for generating sources and energy storage systems, which clearly shows that both elements need to be addressed formally and technically.

Poland’s transmission system operator PSE publishes information on available connection capacity for the transmission grid and updates it at least quarterly. In 2026, PSE also published notices concerning non-market redispatch of PV installations. These system-level data should not, however, be applied directly to a specific industrial installation: the conditions for a plant must be verified with the relevant distribution system operator. PSE data are a system-level signal that grid availability and generation constraints need to be analysed before the project is finalised.

If exports are constrained, the options may include changing the size of the generating source, power control, better matching generation to demand or adding energy storage. Each option should be calculated for the specific load profile.

If one of the investment objectives is also to reduce Poland’s capacity charge or charges for exceeding contracted demand, that effect must be calculated separately using the actual load profile and the current settlement rules. The presence of a PV system alone does not guarantee that these costs will fall.

Roof or ground? Area is only one of the constraints

A large industrial roof looks like a natural location for photovoltaics, but the number of available square metres alone does not determine how much PV should be built. The first step is to check whether the location is technically and operationally suitable.

Photovoltaics on an industrial building roof

For a rooftop installation, relevant factors include:

• structural load capacity and condition;
• the planned service life and timing of the next roof refurbishment;
• areas excluded by rooflights, HVAC equipment, smoke vents and service access routes;
• shading from chimneys, parapets and neighbouring structures;
• the ability to route cables and carry out maintenance safely;
• how the installation will be connected to the plant’s internal electrical network;
• fire safety requirements.

Under Polish law, for photovoltaic equipment with installed electrical capacity greater than 6.5 kW, the design must be agreed with a fire protection expert. After installation is completed, the competent authority of the State Fire Service must also be notified. The current procedures are described by Poland’s National Contact Point for Renewable Energy Sources: design approval and notification of the State Fire Service.

If the roof is due to be refurbished in a few years, the PV system should be assessed together with that refurbishment. Covering the full roof area with modules on a roof that will soon need replacement can create unnecessary removal and reinstallation costs.

Ground-mounted photovoltaics

A ground-mounted system offers more flexibility in module layout but introduces other constraints. These include the distance from the plant’s connection point, cable route, site development plan, possible plant expansion, conflicts with internal logistics and the administrative requirements applicable to the location and scale of the investment.

In practice, the best solution may not be “roof or ground” but a split installation across several areas. Such an option only makes sense if it does not overcomplicate the internal network and still matches the plant’s energy profile.

When does energy storage make sense with industrial photovoltaics?

Energy storage should not be an automatic add-on to a large PV system. Its function should be defined first.

When combined with photovoltaics, storage can, among other things:

• absorb part of the surplus generation around midday and deliver it during later hours;
• help keep grid exports below a specified limit;
• help reduce short-duration demand peaks if the load profile and settlement method justify such operation;
• provide backup power, but only in a system designed for island operation and equipped with appropriate controls.

When assessing storage, power in kW must be distinguished from energy capacity in kWh. The analysis also needs to cover round-trip efficiency, degradation, number of cycles, control strategy, the cost of charging energy and the value of discharged energy.

If the sole objective is to shift surplus PV generation from midday to the evening, the storage profile required is different from one used to reduce short demand peaks. Technology and use cases should therefore be analysed separately in the context of energy storage for industry.

A standard grid-connected PV system does not automatically supply the plant during a grid outage. If resilience to supply interruptions is an investment objective, the backup function must be deliberately designed – with storage or another source, automation, separation of critical loads, protection systems and a permitted method of transitioning to island operation for the specific system.

How should a plant prepare for a PV investment analysis?

Before discussing a specific system capacity, it is worth preparing data that will allow several options to be modelled and apparently attractive solutions to be rejected.

The most useful data package includes:

1. an electricity demand profile covering at least one representative year – hourly, and preferably 15-minute data if available;
2. invoices and information on the tariff, contracted demand and connection conditions;
3. the operating calendar: shifts, weekends, public holidays, shutdowns and production seasonality;
4. a list of planned changes that will affect electricity consumption in the coming years;
5. the single-line diagram and data for transformers, switchgear and main metering points;
6. roof documentation or information on available land;
7. preliminary information from the network operator on connection of the generating source and any export constraints;
8. a definition of whether the energy storage system is intended to solve a specific problem and what that problem is.

This package does not replace a design or detailed capacity sizing. It does, however, allow the investment to be prepared so that subsequent calculations concern realistic options. We describe the detailed sizing process separately in the article how to choose the optimal photovoltaic capacity.

The most common mistakes when preparing an industrial photovoltaic installation

1. Sizing only to annual consumption

Annual energy is a balance, whereas photovoltaics operates at specific times. Without a time-series profile, it is impossible to determine how much generation will be used on site.

2. Designing the system to “fill the entire roof”

The available area indicates the maximum siting potential, not the option best suited to the company. Using the entire roof may be justified, but this must follow from the consumption model, export conditions and economics.

3. Ignoring weekends and shutdowns

This is particularly risky for companies with intensive Monday-to-Friday production. Surpluses often appear precisely when the plant is not operating.

4. Treating storage as a way to fix an oversized PV system

Storage can improve energy utilisation, but it has its own cost, losses and constraints. The first question should be whether changing the PV option or the way loads are controlled would be more effective.

5. Checking grid conditions too late

A concept prepared without the network operator may need to be revised after connection conditions are issued. In large projects, this can affect the schedule, protection systems, controls and costs.

6. Assuming PV will supply the plant during a grid outage

That function requires the right architecture. The mere presence of rooftop modules and grid-tied inverters does not automatically create an emergency power supply.

Industrial photovoltaics should be part of the plant’s energy system

A good industrial photovoltaic investment is not the largest system that can be fitted on the property. It is a system matched to the plant’s load profile, technical, grid and economic conditions, and the company’s development plans.

The first step is therefore to review the hourly load profile, operating calendar, self-consumption, electrical infrastructure and roof, ground and grid constraints. Only then does it make sense to compare specific capacity and technology options.

If a company is preparing a PV investment, analysing these data helps structure the decision before the design stage. In a broader perspective, it is also worth combining this work with electricity cost optimisation and planned changes in the plant’s other energy systems.

FAQ

Is annual electricity consumption enough to assess industrial photovoltaics?

No. Annual consumption shows the scale of demand, but not its timing. Assessing self-consumption and surplus generation requires at least an hourly profile, and for large consumers it is worth using available 15-minute data together with the plant’s operating calendar.

Does a large roof mean that a large PV system is worth building?

Not always. The roof defines potential installation space, but the investment size should account for the demand profile, grid conditions, structural condition, future refurbishment and how surplus electricity will be managed.

Does energy storage always improve the economics of industrial photovoltaics?

No. Storage should have a specific function and be assessed using the actual energy profile. At some plants it may increase the use of PV electricity or limit exports; at others its cost and losses may not be justified.

Why can two plants with the same annual consumption need different PV systems?

Because they may consume electricity at different times of day, on different days of the week and in different seasons. A plant with high demand during daylight hours has different self-consumption potential from a company that uses a large share of its electricity at night or stops production at weekends.

Sources

• PVGIS – hourly solar radiation and PV generation data, Joint Research Centre
• PVGIS – tools for assessing photovoltaic generation, Joint Research Centre
• PSE – information on available connection capacity
• PSE – non-market redispatch of renewable energy sources
• TAURON Dystrybucja – documents for connecting generating sources and storage systems
• Poland’s National Contact Point for Renewable Energy Sources – fire safety requirements for PV installations
• Poland’s National Contact Point for Renewable Energy Sources – State Fire Service notification after installing PV >6.5 kW