Financing from savings – how do ESCO models work?

Financing an energy-efficiency upgrade from future savings is often presented with a simple slogan: “the investment pays for itself”. In practice, however, the most important question is different: how can we credibly determine how much energy the company would have used without the upgrade and what share of the difference is the actual effect of the project?

The ESCO model and performance-based contracts are built around the answer to that question. A well-designed contract defines more than the technology and the payment amount. It should describe the baseline, the method for measuring and verifying savings, the rules for adjustments when production or operating conditions change, the method for converting energy into monetary value, the allocation of risks, and what happens to the equipment after the contract ends.

In short: ESCO financing can reduce the company’s upfront CAPEX and transfer part of the technical risk to the provider. It does not, however, automatically mean that 100% of the project will be financed from savings, that the client bears no risk, or that comparing two energy bills is sufficient.

What is an ESCO and how does it differ from EPC?

ESCO is a commonly used term for an energy service company that may design, implement, finance or operate a project that improves energy efficiency. The acronym alone does not determine who provides the capital or who owns the equipment.

EPC (Energy Performance Contracting), by contrast, refers to a contract in which settlement is linked to an agreed energy performance result or another performance parameter, and the achieved results are monitored and verified during the contract term. In practice, one ESCO may offer several EPC variants, while some service projects may use a similar performance-based settlement mechanism without having an identical financing structure.

The most common mechanisms can be arranged as follows:

Commercial names vary. From the company’s perspective, the label “ESCO” matters less than the answers to four questions: who finances the project, who owns the asset, who is responsible for performance, and which method will be used to calculate that performance.

How does financing an upgrade from future savings work?

A well-prepared project can be broken down into seven consecutive stages:

1. Collect data and establish the baseline. Define the technical condition, historical energy use and the factors that affect it: production volume, operating hours, outdoor temperature, number of shifts, process parameters or occupied floor area.
2. Select the scope of the upgrade. The technical analysis should identify which measures deliver a predictable and measurable result and what their investment, service and financing costs are.
3. Define the guarantee level and M&V rules. The parties specify which savings are guaranteed, in which units, how often they will be verified and which adjustments are permitted.
4. Structure the financing. Capital may be provided by the ESCO, a bank, a leasing company, an external investor or the company itself. A blended structure is also possible.
5. Implementation, commissioning and acceptance. It is necessary to confirm not only completion of the works but also equipment operating parameters, setpoints, metering and the system’s readiness to collect data.
6. Report and settle savings. Actual data are compared with an appropriately adjusted baseline, and the result becomes the basis for settlement under the contract.
7. Close the contract. Final settlement takes place and, where provided for in the contract, ownership of the equipment, energy management system, documentation, licences and data is transferred.

This mechanism differs from a conventional loan primarily because energy performance becomes part of the contract and affects settlements between the parties.

Baseline: why last year’s bill is not enough

Under Poland’s Energy Efficiency Act, energy savings are defined as the difference between energy potentially consumed before a measure and energy consumed after its implementation, taking normalised external conditions into account. This illustrates the core issue well: savings are a comparison with a situation that no longer exists after the upgrade.

Therefore, the baseline should not automatically be equated with a single bill or the sum of bills from the previous 12 months. It is necessary to determine whether the baseline period was representative.

For example, it is worth checking:

• production volume and product mix,
• number of operating hours and shifts,
• maintenance shutdowns and failures,
• outdoor temperature and heating or cooling demand,
• pressure levels, temperatures and other process setpoints,
• building occupancy or number of users,
• start-up or shutdown of production lines,
• expansion of the plant’s floor area,
• a change in the energy source or the introduction of on-site generation.

Depending on the facility, a better basis may be the most recent typical year, several typical years, or a model based on hourly data and explanatory variables. The key is to agree the method before the settlement period begins.

In simplified form, the mechanism can be expressed as follows:

Energy savings = adjusted baseline consumption − consumption during the reporting period ± agreed non-routine adjustments.

Auditor’s comment: if the baseline is weak, even a technically very good upgrade may later lead to a dispute over the result. In an ESCO project, the quality of pre-investment data is therefore part of the financial model, not merely an addition to the audit.

How a change in production can completely alter the savings picture

Assume that before the upgrade the plant used 10 000 MWh per year while producing 100 000 units, i.e. an average of 0.1 MWh per unit.

After the upgrade, production increased to 120 000 units and energy use was 10 500 MWh. A simple comparison of bills would show an increase in consumption of 500 MWh and could suggest that there were no savings.

However, if we assume solely for illustration that the relationship between energy use and production is linear and all other conditions remained unchanged, the adjusted baseline for 120 000 units would be:

120 000 × 0.1 MWh = 12 000 MWh.

In this simplified example, the savings are therefore:

12 000 MWh − 10 500 MWh = 1 500 MWh, or 12.5% relative to the adjusted baseline consumption.

At the same time, the energy intensity of production falls from 0.1 to 0.0875 MWh per unit.

This is a hypothetical example, not a ready-made method for every plant. In a real process, the relationship may be non-linear: part of consumption is fixed, part depends on production, and part on temperature, product mix or capacity utilisation. The model should therefore be based on data from the specific company.

Weather, operating hours and other factors requiring adjustments

Weather conditions

For heating, cooling and ventilation projects, outdoor temperatures can have a major impact on consumption. A warmer winter can reduce bills even without an upgrade, while a colder one can increase consumption despite a correctly functioning project.

The contract should therefore define, among other things, the source of weather data, how they are assigned to the facility and the adjustment method. Depending on the project, degree days or a statistical model describing the relationship between consumption and temperature may be used.

Operating hours and load

A change from two shifts to three, weekend operation, commissioning an additional line or shortening production time may affect consumption more than the upgrade itself. These parameters should be recorded in a way that allows them to be used in the settlement model.

Process parameters

In a compressed air system, for example, an increase in the required pressure may affect the result. For a refrigeration system, a change in evaporation temperature or the required product temperature may be relevant. In ventilation, changes in airflow rate and operating schedule matter.

Non-routine events

Changes that cannot be treated as normal variability require a separate procedure: expansion of a production hall, replacement of a large machine unrelated to the ESCO project, closure of a department, a change in production technology or an extended shutdown.

A good contract describes not only the adjustment formula but also who must report such an event, by when, what data must be provided and how a failure to reach agreement is resolved.

Energy prices: separate technical savings from monetary savings

In a multi-year contract, the price of energy may change many times. A price change alone does not mean that the upgrade has started saving more or less energy.

It is therefore worth separating two levels:

• technical savings, e.g. MWh of electricity, GJ of heat or m³ of gas;
• financial savings, i.e. the value of the technical effect after applying the agreed price, tariff and cost scope.

Let us return to the hypothetical saving of 1 500 MWh. At a reference price of 500 PLN/MWh, its value is 750 000 PLN. At a price of 650 PLN/MWh, the same technical saving corresponds to 975 000 PLN in avoided cost.

It is the same energy saving, but with a different monetary value.

The contract should therefore clearly specify:

• whether settlement uses a reference, actual or indexed price,
• which invoice components are included,
• how changes in tariffs and the structure of charges are treated,
• what happens when the energy supplier changes,
• whether the ESCO’s remuneration depends on MWh, PLN or both,
• how a benefit resulting solely from a price change is settled.

If the provider is primarily responsible for the technical performance of the upgrade, transferring the full risk of future energy prices to it would usually mix two different types of risk.

How to measure and verify savings

Measurement and verification, or M&V (Measurement & Verification), answers the question of how data can be used to demonstrate the project’s performance. IPMVP emphasises that savings cannot be measured directly because they represent avoided consumption. The conditions before and after the upgrade must be compared on a consistent basis, with appropriate adjustments.

The M&V plan should be agreed before settlement begins and should include at least:

1. the project measurement boundary,
2. the baseline period and reporting periods,
3. meters, sensors and data sources,
4. independent variables affecting consumption,
5. formulas and models used for adjustments,
6. the procedure for missing or erroneous data,
7. a list of events requiring non-routine adjustments,
8. reporting frequency,
9. the report approval process,
10. data audit rules and dispute resolution procedures,
11. accuracy or uncertainty requirements, where material to the value of the contract.

It is not always necessary to measure the entire facility. For a simple measure, the measurement boundary can be set around the upgraded equipment. For a package of multiple interacting measures, whole-facility settlement may be more appropriate. Where baseline data cannot be obtained reliably, a calibrated model may be used. The selected method should reflect the scale of the project and the financial risk rather than becoming an end in itself.

A savings guarantee is not the same as a forecast

A forecast states the expected result. A guarantee defines the minimum performance for which the provider accepts responsibility under the conditions set out in the contract.

This distinction is crucial. If, for example, 1 500 MWh per year is guaranteed, the contract must also explain:

• which baseline the guarantee applies to,
• which adjustments may be applied,
• what tolerance or uncertainty level applies,
• which operating obligations the company has,
• which events exclude liability,
• what happens if there is a savings shortfall,
• how savings above the guarantee are shared.

Different mechanisms are possible. In one model, the ESCO financially covers any shortfall below the guaranteed level. In another, its remuneration is reduced. Under a shared-savings model, the parties may divide the verified benefit in an agreed proportion.

Example: if verified financial savings amount to 750 000 PLN and, under the contract, 500 000 PLN is allocated to remuneration and project financing service, 250 000 PLN remains with the company. If the result is 600 000 PLN or 900 000 PLN, the way the shortfall or surplus is settled should follow from the contract – not from negotiations after the fact.

How benefits and risks can be allocated

There is no universal rule that “under an ESCO model the client bears no risk”. A rational contract assigns each specific risk to the party that has a genuine ability to manage it.

Risk should be allocated deliberately. Attempting to transfer every uncertainty to the ESCO usually does not remove it from the project – it may simply increase the price, the number of exclusions or the difficulty of obtaining financing.

Who finances the equipment in an ESCO model?

Contrary to a common simplification, an ESCO does not always have to finance 100% of the investment expenditure from its own capital. The structure may take different forms:

• the ESCO finances the investment from its own funds,
• the ESCO obtains a loan or other financing and includes its cost in its remuneration,
• the client finances the investment, while the ESCO is responsible for design, implementation and the guaranteed result,
• capital is provided by a bank, leasing company, fund or another external investor,
• part of the investment expenditure is financed by the client and part by the ESCO or financier,
• the project combines an EPC model with available public support.

Official Polish EPC guidelines explicitly point out that assuming 100% of the cost of every project will be financed solely from savings may be incorrect. In addition to CAPEX, the cost of financing, energy management, service and maintenance, as well as the margin of the provider and financier, must be covered.

Before comparing offers, it is therefore worth calculating not only the monthly payment but also the full cost of the contract over its entire term and the value of the risk actually assumed by the provider.

Who owns the equipment?

There is no single model in this area either.

The equipment may:

• belong to the company from the outset,
• remain the property of the ESCO during the contract term and later be transferred to the client,
• belong to a leasing company or another financier,
• remain an asset of the provider even after the end of the basic contract term if the service is structured that way.

The contract should therefore specify at least:

• when ownership transfers,
• the risk of accidental loss or damage,
• the obligation to insure the equipment,
• the rules for repair and replacement,
• the ESCO’s access to the equipment,
• the financier’s ability to establish security interests,
• the buyout or settlement value in the event of early termination,
• the required technical condition upon handover,
• rights to software, data and documentation.

This is important not only from a legal perspective. Ownership and the payment structure may affect the accounting treatment of the contract. A private company should therefore not assume that it will automatically obtain “off-balance-sheet financing” or that the contract will have no effect on its financing capacity. Such consequences must be assessed for the specific contract and the accounting principles applied by the company.

Polish regulations provide specific conditions concerning the impact of EPC on the debt and deficit of public-sector entities. This is not, however, a universal rule that can be applied directly to the balance sheet and borrowing capacity of a private company.

What happens when the ESCO contract ends?

The end of the contract should be designed at the time the agreement is signed.

Depending on the project structure, the parties should determine:

• final settlement of guaranteed savings,
• the date and method of transferring ownership of the assets,
• the required condition of the equipment and how wear is assessed,
• transfer of manufacturers’ warranties and service agreements,
• handover of as-built documentation and service history,
• handover of measurement data and baseline history,
• transfer of the energy management system,
• rights to software, licences and administrator accounts,
• handover of instructions, passwords, configurations and backups,
• training of the technical team,
• availability of spare parts and support after the contract ends.

The early termination formula is particularly important. It should define how the settlement amount is calculated, the value of the remaining financing, any residual value of the assets, and the parties’ rights to the equipment and data. Without such a mechanism, a project that has worked well for several years may become difficult to unwind when the plant is sold, technology changes or the company is reorganised.

Can ESCO be combined with Polish white certificates or other support?

An ESCO model may form part of a broader project financing structure. This does not mean, however, that every form of support can automatically be added to every contract.

For Polish white certificates, it is necessary to assess separately the eligibility of the measure, when the work starts, which entity is entitled to submit the application, and which party, under the contract, will receive the rights and economic benefits arising from the energy efficiency certificates. We explain the mechanism itself in more detail in our article on Poland’s white certificate system.

Similarly, for grants and preferential financing instruments, the rules of the specific programme and the impact of the support on the contract structure must be verified. We discuss this topic in more detail in our materials on energy-efficiency grants.

The key principle is simple: public support and the ESCO model are two separate elements that must be combined deliberately rather than treated as automatically cumulative sources of benefits.

How to assess a company implementing an ESCO project

A request for proposal should not begin solely with questions about the contract term and the monthly fee.

It is worth asking a potential provider to present:

• the technical scope and assumptions used to select the technology,
• the data used to establish the baseline,
• the measurement and verification method,
• the level of guaranteed savings and all exclusions,
• the method for adjustments when production and other conditions change,
• the financing structure and cost of capital,
• the allocation of risks among the client, the ESCO and the financier,
• the service scope and responsibility for equipment availability,
• the asset ownership model,
• the rules for settling savings shortfalls and surpluses,
• the procedure for ending and terminating the contract early,
• the client’s access to raw measurement data.

Transparency of calculations is particularly important. The client should be able to trace the path from the raw meter reading, through baseline adjustments, to the value shown in the ESCO settlement.

Questions worth asking before signing an ESCO contract

Before making a decision, the company should know the answers to at least the following questions:

1. What exactly is guaranteed: MWh, a percentage of savings, PLN, equipment availability, or several parameters at the same time?
2. Which period was selected as the baseline and why was it considered representative?
3. Which variables will be normalised: production, temperature, operating time, floor area, number of users, process parameters?
4. What is the baseline adjustment formula and can it be reproduced independently?
5. How will major non-routine changes in the company be accounted for?
6. Which meters and data form the basis for settlement, and who is responsible for their legal compliance, completeness and archiving?
7. Who has access to the source data?
8. How are technical energy savings converted into PLN?
9. How does the contract treat changes in energy prices, tariffs and supplier?
10. What happens if the guaranteed performance is not achieved?
11. How are savings above the guaranteed level shared?
12. Who finances the CAPEX and what is the total cost of that financing?
13. Who owns the equipment in each year of the contract?
14. Who pays for service, parts, failures and replacement of equipment?
15. Can the client change how the installation is operated, and what are the consequences for the guarantee?
16. What is the amount and procedure for early termination of the contract?
17. What exactly does the client take over at the end: equipment, documentation, data, licences, controllers and administrator access?
18. How are disputes over the calculation of savings resolved?
19. Is independent verification of the model and M&V reports available?
20. How will the contract operate after plant expansion, a change in the production profile or the sale of the company?

If the answers to these questions cannot be found in the contract or technical appendices, the risk is usually greater than the financial presentation alone suggests.

When can conventional investment financing be simpler or less expensive?

ESCO delivers the greatest value when a company wants to combine capital, technical responsibility, service and a performance guarantee in one multi-year model. It is not always the least expensive solution, however.

Conventional financing may be more advantageous when:

• the company has access to low-cost capital or sufficient own funds,
• the measure is technically simple and has a short payback period,
• the result can be measured easily without an extensive M&V system,
• the company has its own team capable of managing the project and maintenance,
• the company wants to own the equipment from the outset and retain full freedom to upgrade it,
• the cost of the ESCO margin, financing and multi-year service exceeds the value of the risk transferred,
• the project is likely to undergo significant changes over the next few years,
• operating conditions are so unstable that a long-term guarantee would require numerous complex adjustments.

The comparison should therefore not be reduced to the question “do we need CAPEX upfront?”. It is worth comparing the total cost of ownership, financing cost, value of the guarantee, M&V and service costs, and the risks that remain with the company.

If the objective is to compare ESCO with other financing methods, see our separate guide: how to finance energy-efficiency upgrades in a company. Here, we deliberately focus on the mechanics of settling a project from future savings.

How to prepare a company for discussions about ESCO financing

The better the data are prepared, the easier it is to obtain comparable offers and avoid disputes over the baseline.

Before starting discussions, it is worth collecting:

• energy bills and consumption data covering the full seasonal cycle,
• hourly or 15-minute data, if available,
• production volume and product-mix data,
• equipment operating schedules and production-shift schedules,
• information on shutdowns, maintenance and failures,
• process parameters affecting energy use,
• a history of changes at the plant,
• information on planned expansions and new lines,
• an inventory of the main energy end uses and energy sources,
• available local measurements,
• weather data, where relevant,
• process-quality and comfort requirements,
• existing service agreements and technical constraints.

It is also worth defining in advance which measures are to be covered by the guarantee. A simple lighting upgrade will require a different measurement structure from a package covering a heat source, controls, ventilation and several production processes.

The key takeaway

Financing from savings works well not because “the bill after the upgrade is lower”. It works when the parties can define a credible baseline, measure performance on consistent terms and assign each risk to the party that can actually manage it.

A well-designed ESCO model can reduce the barrier of upfront CAPEX, combine the upgrade with service and transfer an agreed share of technical risk to the provider. A poorly designed contract, however, can turn a straightforward investment into a multi-year dispute over whether the savings result from the equipment, weather, production or the energy price.

Before negotiating how future savings will be shared, it is therefore worth accurately calculating the energy performance itself. Only then can the company assess whether the ESCO model genuinely offers an advantage over conventional investment financing.

Sources and methodological basis

• Act of 20 May 2016 on Energy Efficiency (Poland) – consolidated text, Journal of Laws 2025, item 711: https://eli.gov.pl/eli/DU/2025/711/ogl
• Polish Ministry of Climate and Environment, “Guidelines for Energy Performance Contracts (EPC)”: https://www.gov.pl/web/klimat/wytyczne-do-umow-o-poprawe-efektywnosci-energetycznej-epc
• Directive (EU) 2023/1791 of the European Parliament and of the Council on energy efficiency: https://eur-lex.europa.eu/legal-content/PL/TXT/?uri=CELEX:32023L1791
• Commission Recommendation (EU) 2024/2476 on energy services: https://eur-lex.europa.eu/legal-content/PL/TXT/?uri=CELEX:32024H2476
• Efficiency Valuation Organization, IPMVP: https://evo-world.org/en/products-services-mainmenu-en/protocols/ipmvp

Sources and legislation verified as at 19 August 2026. Before signing a specific contract, its legal, tax, accounting and financial implications for the company concerned should be assessed separately.