5 Ways to Interpret PVSyst Self Consumption Results

When evaluating self-consumption solar power generation in PVSyst, the results for Self Consumption become more important than the generation amount itself. For a full-feed-in model, you can grasp the overall direction by focusing on annual energy production, PR, losses, and the amount of electricity sold. However, for self-consumption systems you need to look at how much of the generated electricity was used on-site, how much surplus power was produced, how much purchased electricity was reduced, and at which times the load curve and the generation curve overlap.

Self Consumption is translated into Japanese as "自家消費". In PVSyst's results, the electricity generated by the PV is categorized into the amount used by the load, the surplus energy exported to the grid, the electricity purchased from the grid, and, where applicable, the amounts charged to and discharged from the battery. If this is not read correctly, you can make incorrect judgments — for example, you may find that despite large generation the economic benefit is not as great as expected, or conversely that even with modest generation the reduction in purchased electricity is large.

In this article, I organize five perspectives you should check when reading PVSyst's Self Consumption results, arranged in an order that is practical for everyday use. The focus is on on-site solar PV systems used at factories, warehouses, commercial facilities, schools, hospitals, public facilities, and residences — not off-site systems. Specific screen names and labels in PVSyst may vary somewhat depending on the version and settings, but the basic way of reading them is the same.

Table of Contents

• Self Consumption measures the amount of generated electricity that was actually used, not the total generation

• First, read the annual self-consumption rate and the self-sufficiency rate separately

• Look at the monthly results to see seasonal differences in surplus and purchased electricity

• Use time-of-day curves to examine the overlap between load and generation

• For results with a battery, read the charge/discharge losses and the remaining charge

• Summary

Self Consumption measures the amount of electricity actually used, not the amount generated.

When reading PVSyst's Self Consumption results, the first thing to be aware of is that high generation does not equal a large self-consumption effect. In photovoltaic simulations, the larger the annual generation, the better the result tends to appear. However, for self-consumption systems, if the generated power cannot be used on-site, it will either flow to the grid as surplus, be curtailed, or be temporarily stored in a battery.

For example, even if a system generates 1,000,000 kWh per year, if only 500,000 kWh of that can be used on-site, the direct reduction in purchased electricity from self-consumption is 500,000 kWh. Conversely, even if annual generation is 800,000 kWh, if the system is designed so that 700,000 kWh can be used on-site, the reduction in purchased electricity may be greater in the latter case. It is important to discern this difference in PVSyst's Self Consumption results.

The main flows to look at in Self Consumption are PV generation, load, direct self-consumption, surplus, and purchases from the grid. PV generation is the amount of electricity produced by solar power. Load is the amount of electricity consumed by the facility. Direct self-consumption is the portion of PV generation that was used by the on-site load. Surplus is the electricity that was generated but could not be used at that moment. Purchases from the grid are the amount of electricity received from the utility when PV alone is not sufficient.

What should be noted here is that if you look only at annual values, time shifts are hard to see. Even if the annual load is 1,000,000 kWh and annual PV generation is also 1,000,000 kWh, in facilities that generate during the day and have large loads at night, not all of it can be self-consumed. PVSyst matches generation and load on a time-step basis, so the Self Consumption results reflect the effects of this time mismatch.

Therefore, when interpreting the Self Consumption results, the first thing to check is not the annual generation but where the generated electricity flowed. Separate and view whether PV generation was used directly by the load, was surplus, went into the battery, or was sent to the grid. By examining this breakdown, you can tell whether the system capacity is too large, whether it matches the load data, whether a battery is necessary, and whether there is room on the operations side to increase daytime load.

For self-consumption solar systems, simply increasing the panel capacity is not always the best approach. Up to a certain capacity it effectively reduces purchased electricity, but beyond that surplus increases and the self-consumption rate can decline. The Self Consumption results in PVSyst are a resource for finding that boundary. By reading annual generation, direct self-consumption, surplus energy, and purchased electricity side by side, it becomes easier to judge the appropriateness of the system capacity.

First, read the annual self-consumption rate and the self-sufficiency rate separately

In Self Consumption results, the terms self-consumption rate and self-sufficiency rate are often confused. These two are similar, but their meanings are different. In PVSyst’s notation, the names may vary depending on project settings and output items, but conceptually you must always read them separately.

The self-consumption rate indicates the proportion of the electricity generated by PV that was used within the facility. It is a ratio seen from the generation side. For example, if a PV system generates 1,000,000 kWh and 700,000 kWh of that is used within the facility, the self-consumption rate is 70%. The remaining 300,000 kWh becomes surplus, electricity sold to the grid, output curtailment, battery-related flows, and so on.

On the other hand, the self-sufficiency rate is an indicator that shows how much of a facility's electricity consumption was covered by PV. It is the proportion seen from the load side. For example, if a facility's annual electricity consumption is 2,000,000 kWh and 700,000 kWh can be supplied from PV either directly or via batteries, the self-sufficiency rate is 35%. In other words, the self-consumption rate indicates how much of the generated electricity was used up, while the self-sufficiency rate indicates how much of the electricity consumed was replaced by PV.

If you look at the results without understanding this difference, you will make an incorrect assessment. A high self-consumption rate does not necessarily mean a high reduction in purchased electricity for the entire facility. When a small PV system is connected to a large load, almost all of the generated power can be used, so the self-consumption rate becomes high. However, if the PV capacity is small, its contribution to the facility’s total consumption is limited, and the self-sufficiency rate remains low.

Conversely, installing a large PV system can increase the self-sufficiency rate, but the self-consumption rate may decrease because the amount of power that cannot be used during the daytime rises. If a large-capacity PV array is installed on a factory roof, much of the output may be consumed during weekday daytime hours, but surpluses can increase during weekends and long holidays. In this case, not only the annual generation but also holiday load, lunch breaks, and seasonal operating conditions greatly affect the Self Consumption results.

In PVSyst’s annual results, first check PV generation, load consumption, self-consumed energy, surplus energy, and grid purchases. On that basis, determine which is higher and which is lower between the self-consumption rate and the self-sufficiency rate. If the self-consumption rate is high and the self-sufficiency rate is low, the PV capacity is modest, surplus is small, but there may still be room to reduce purchased electricity. If the self-consumption rate is low and the self-sufficiency rate is high, the PV capacity is large; reductions in purchased electricity proceed, but addressing surplus may become a challenge.

In practice, these two indicators are considered simultaneously to decide which one to prioritize for a given project. If selling surplus electricity is possible and a certain level of unit price can be expected, there may be a decision to increase system capacity even if the self-consumption rate declines somewhat. On the other hand, when reverse power flow is not permitted, the unit price for surplus electricity is low, output curtailment is to be avoided, or there are constraints in the power supply contract, a design that keeps the self-consumption rate high is required.

When reading Self Consumption results, do not simply judge “it’s good because the self-consumption rate is X%”; instead, consider the load size, PV capacity, how surplus is handled, the conditions for selling electricity, the electricity tariff, contracted power, and the customer's operations together. The PVSyst figures are used as material for that judgment.

Examine seasonal differences between surplus and purchased electricity in monthly results

After checking the annual results, next we look at the monthly Self Consumption results. In self-consumption solar, annual values alone are often insufficient for assessment. In particular, solar generation varies by season, and facility loads also change seasonally, so the differences between months can be large.

In spring and autumn, solar radiation is relatively good, and surplus generation tends to occur at facilities with low air-conditioning loads. In summer, while generation is also high, facilities with large cooling loads tend to increase self-consumption. In winter, solar radiation decreases depending on the region, and effects from snowfall and low solar altitude appear, but facilities with heating or production loads may still see reductions in purchased electricity. Reading PVSyst’s monthly results reveals this seasonal relationship between generation and load.

First, in the monthly results you should check which months the surplus energy is concentrated in. Whether the surplus is large only in specific months or persists throughout the year will change the countermeasures. If the surplus is large only in spring, the impact on annual economic performance may be limited. On the other hand, if surpluses are large in almost every month, possible causes include PV capacity being too large relative to the load, underestimated load data, holiday settings not being reflected, or the need to consider batteries and control strategies.

The next thing to check is which months still have electricity purchases from the grid. Even after installing PV, purchases from the grid are necessary at night and during bad weather. In particular, large winter grid purchases mean that solar alone cannot cover seasonal demand. Large summer grid purchases suggest that cooling or production loads may be exceeding PV generation. Reviewing monthly grid purchases lets you determine whether to increase PV capacity, consider load shifting, or install battery storage.

In the monthly Self Consumption results, you cannot simply say that months with higher generation are better. Even if generation is high, if there is a lot of surplus the self-consumption effect is limited. Conversely, in months with low generation, if it aligns well with the load, that month’s generation may be almost entirely effective at reducing purchased electricity. For self-consumption systems, it is important to consider the amount of generation together with how it is used.

Monthly results also help verify the validity of the load data. For example, if a factory is actually supposed to operate at high capacity in summer but the summer load in PVSyst is set low, the surplus may appear excessive. Conversely, if holidays are set to the same load as weekdays, self-consumption may be overestimated. If the Self Consumption results seem off, you should check the load data assumptions before the PV model.

When evaluating self-consumption in PVSyst, a simple load model that only inputs annual energy consumption may not adequately reflect actual variations by time of day and day of week. If possible, it is desirable to use 30-minute or 1-hour power consumption data to create a load profile that reflects weekdays, holidays, and seasonal variations. Monthly results also serve as checkpoints to verify whether that input data is close to reality.

Viewing the Overlap of Load and Generation in Time-of-Day Curves

To read the Self Consumption results in depth, you need to look at the time-of-day curves. Annual tables and monthly tables are useful for grasping the overall picture, but the essence of self-consumption is temporal coincidence. If there is load during the time when PV generation occurs, it will be self-consumed; if there is no load, it becomes surplus. If there is load during a period when there is no PV generation, power is purchased from the grid.

One common way to check is to overlay the daily generation curve and the load curve. On a sunny day, PV generation increases in a bell-shaped curve during the daytime, and if the facility load falls below it, a surplus will occur. If the facility load is larger than PV generation, the generated power is mostly consumed on-site and any shortfall is purchased from the grid. Morning, evening, and nighttime loads cannot be met by PV alone, so electricity is purchased from the grid.

When reading this curve, check when peaks in the load occur. In factories where production equipment and air conditioning run during the daytime, loads tend to overlap with PV generation, making them suitable for self-consumption. In factories with significant nighttime operations, refrigerated warehouses, lodging facilities, and residences, part of the load occurs outside PV generation hours, so the direct self-consumption rate tends to be low. In commercial facilities, operating hours often overlap with generation hours, but they are affected by closed days and seasonal variations.

What you should pay attention to in time-of-day curves is the surplus around noon. Because solar power tends to peak around midday, a low load during this period leads to a large surplus. Conversely, in facilities where load is high in the morning and afternoon but drops only at midday, lunch breaks or process stoppages can be the cause of the surplus. If PVSyst results show a noticeable midday surplus, options include reducing system capacity, considering load shifting, installing batteries, or implementing controls.

Also, the holiday curve is important. Even if weekday self-consumption rates are high, facilities with low loads on Saturdays, Sundays, and public holidays can see surpluses concentrated on holidays. Even if this does not look very large in annual values, for projects where reverse power flow is not allowed, instantaneous surpluses on holidays can become a problem. When reading PVSyst's Self Consumption results, it is necessary to check not only representative weekdays but also holidays, extended vacations, and low-load days.

In time-of-day analysis, the peak-cut effect and the energy reduction effect are considered separately. If PV is generating during daytime peak demand hours, it may affect contracted power and demand. However, if PVSyst’s Self Consumption results are presented mainly as annual energy, the certainty of demand reduction cannot be judged simply. Because PV output drops on cloudy or rainy days, if you expect to reduce contracted power you need to carefully check actual demand records and weather conditions separately.

Time-of-day curves can also be used to optimize system capacity. By gradually increasing PV capacity and observing at which time periods surplus begins to appear, you can identify capacity ranges that are economically advantageous. While the initial several tens of kW can be almost entirely self-consumed, as capacity increases daytime surplus grows and the marginal effect of additional panels may decline. Determining this limit is an important point in self-consumption-oriented design.

For results that include a battery, read the charge/discharge losses and the remaining capacity.

When considering Self Consumption in PVSyst, cases where a battery is combined are also common. Installing a battery allows you to temporarily store excess PV power generated during the daytime and use it for loads in the evening or at night. As a result, power that would otherwise have been surplus under direct self-consumption can be utilized, potentially increasing the self-consumption rate and the self-sufficiency rate.

However, adding a battery does not necessarily guarantee better results. Batteries have charging losses, discharging losses, conversion losses, capacity limits, output limits, and upper and lower SOC limits. In PVSyst results, it is necessary to distinguish the energy that entered the battery from the PV, the energy that left the battery to the load, the energy lost in charging and discharging, the surplus that could not be accepted when fully charged, and the periods when the load could not be supplied due to insufficient remaining charge.

First, what I want to confirm is how much the battery is actually being used. Even if you increase capacity, if the amount of charging and discharging is small, the effectiveness relative to the investment will be limited. For example, at facilities with little PV surplus, the power available to charge the battery itself is small. In such cases, increasing battery capacity only increases the amount of unused capacity. Conversely, at facilities with a large surplus and significant nighttime loads, the battery may be able to operate effectively.

The next thing to look at is charge/discharge losses. Power routed through the battery incurs losses compared with direct self-consumption. If PV is used directly to serve loads during the day, losses remain relatively small, but if it is first stored in the battery and then used, the amount of usable energy is reduced by the round-trip efficiency. Therefore, even if a battery reduces surplus, you need to check how much the total usable energy has actually increased.

Also, the SOC trend is important. SOC indicates the battery's state of charge. If it reaches full charge quickly during the day, the capacity may be insufficient and unable to absorb the surplus. Conversely, if it rarely becomes fully charged and the SOC remains low, there may be little PV surplus, the load may be large, the charging conditions may be restrictive, or the capacity may be oversized and not being used up. By observing the SOC trend, you can determine whether the capacity is too small, too large, or appropriate for the operation.

In the Self Consumption results with batteries, it is important not to focus too much solely on the increase in the self-consumption rate. Even if the self-consumption rate rises, when considering battery cost, replacement costs, degradation, maintenance, PCS capacity, installation space, and safety measures, it may not be economically viable. PVSyst’s results indicate how much power can be moved technically, and separate cost calculations are required for investment decisions.

Furthermore, how you evaluate the battery changes depending on whether it is used for peak shaving, surplus absorption, or as an emergency power supply. If it is for surplus absorption, the amount of surplus reduction and nighttime usage are important. If it is for peak shaving, it is important whether the battery can discharge during peak demand periods. If it is for emergency purposes, rather than designing it to be fully discharged during normal times, control to ensure a certain remaining charge may be necessary. When looking at PVSyst's Self Consumption results, you need to interpret them with the battery's purpose clearly defined.

Verify any anomalies in the results using the load data and control conditions

When reading PVSyst's Self Consumption results, if the generation or self-consumption figures seem off, the first things to check are the load data and the control conditions. It's easy to focus on the PV-side settings, but in self-consumption systems the accuracy of the load data greatly influences the results.

Particularly important in load data are time resolution, day-of-week patterns, seasonal variations, holidays, and nighttime load. If you create an average load based only on annual energy consumption, it can make self-consumption appear easier than it actually is. In reality, some facilities have large loads during weekday daytime hours and almost no load on holidays. Conversely, for facilities that operate 24 hours, nighttime loads are substantial, and daytime-only PV may struggle to raise the self-sufficiency rate.

For example, even facilities with the same annual energy consumption can have completely different self-consumption results depending on whether the load is daytime or nighttime. With a daytime load, PV generation overlaps with demand, increasing direct self-consumption. With a nighttime load, PV generation is out of phase, so without batteries the self-consumption rate and the self-sufficiency rate are unlikely to improve. This difference cannot be represented by annual kWh alone.

Control settings are also important. Results vary depending on whether reverse power flow is allowed or prohibited, whether surplus is sold to the grid, output is curtailed, or batteries are charged. In projects where reverse flow is not allowed, PV output must be curtailed during periods of surplus, so there will be a difference between the simulated potential generation and the energy actually available for use. It is important to check how PVSyst treats surplus and unused energy in its results.

Also check the relationship between PV capacity, PCS capacity, and grid interconnection capacity. If PV module capacity is large and PCS capacity is small, clipping may occur at the peak during sunny conditions. In self-consumption systems, designs may reduce PCS capacity to lower costs while matching output to the load, but you must consider the relationship with generation losses and surplus. When viewing Self Consumption results, it is necessary to distinguish between losses where the PV could not generate and surplus where it generated but could not be used.

A typical sequence of checks when results seem questionable is: first, check whether the annual total of the load data matches the actual results. Next, verify that the monthly consumption aligns with the actual electricity usage patterns. Then look at differences between weekdays and holidays, between daytime and nighttime, and between seasons. Finally, check reverse power flow, power exported to the grid, output curtailment, battery control, and PCS capacity settings. Viewing them in this order makes it easier to isolate the cause.

PVSyst's Self Consumption results produce outputs that directly reflect the input conditions. In other words, if the input load deviates from reality, the results will also deviate from reality. Especially in economic assessments of self-consumption systems, it is not an exaggeration to say that the quality of the load data determines the reliability of the simulation. It is desirable to use not only monthly bills but, if possible, smart meters, BEMS, demand monitoring devices, and 30-minute interval data at the point of connection.

Understanding site conditions also affects how results should be interpreted. PV power generation changes depending on the orientation and tilt of roof-mounted or ground-mounted installations, shading, surrounding buildings, snowfall, soiling, and maintainability. Furthermore, if on-site equipment positioning or surveying accuracy is insufficient, discrepancies can arise in shading assumptions and layout planning. Using a system that records site positions by combining an iPhone with high-precision GNSS, such as LRTK, makes it easier to streamline verification of PV equipment layouts, field survey notes, and cross-checking with point clouds and drawings. Connecting PVSyst desktop assessments with on-site information also makes it easier to verify the assumptions behind Self Consumption results.

Summary

When reading PVSyst's Self Consumption results, it's important not to judge solely by the amount of generation. For self-consumption systems, what matters is how much of the generated electricity was used on-site, how much surplus was produced, and how much purchased power from the grid was reduced.

First, check the flow of PV generation, self-consumed electricity, surplus electricity, and purchased electricity. Next, read the self-consumption rate and the self-sufficiency rate separately. The self-consumption rate indicates how much of the generated electricity was actually used, and the self-sufficiency rate shows how much of the facility’s electricity consumption was covered by PV. These two are similar but have different meanings, so it is important not to confuse them.

Then, review the monthly results to check in which seasons surpluses and electricity purchases are concentrated. Even if the annual values appear unproblematic, issues can become apparent when viewed by month or day, such as surplus in spring, electricity purchases in winter, or surplus on holidays. Furthermore, by looking at time-of-day curves you can see when generation and load overlap. The essence of self-consumption is not the total energy amount but whether generation and consumption are aligned in time.

When adding a battery, check not only the increase in self-consumption rate but also charge/discharge losses, the SOC trajectory, how the capacity is utilized, and the investment effect. Batteries can be effective for utilizing surplus, but their effectiveness is limited if the capacity is either oversized or undersized. In PVSyst results, it is important to read how much the battery is actually charging and discharging.

Finally, when the results feel off, check not only the PV-side settings but also the load data and control conditions. Annual energy consumption, monthly consumption, weekday/weekend patterns, daytime/nighttime loads, reverse power flow, electricity sales, output curtailment, PCS capacity, and battery control all have a major impact on the results. Because self-consumption strongly depends on the input assumptions, the more realistic the load data you use, the more useful the results will be for decision-making.

PVSyst's Self Consumption results are the starting point for capacity sizing, surplus mitigation, battery considerations, and economic assessments for self-consumption solar power. Rather than looking only at annual generation, reading in sequence when, where, and how much of the generated electricity was used makes it easier to concretely assess the appropriateness of equipment capacity and identify specific points for improvement.