Techno-Economic Evaluation and Project Decision-Making

Introduction and context

Energy-management decisions are engineering decisions with financial, operational, environmental, and strategic consequences. A technically attractive efficiency measure may be rejected because its cash flow is uncertain, its downtime requirement is unacceptable, or its capital competes with a more valuable project. Conversely, a project with a moderate energy saving may be preferred when it reduces production risk, improves power quality, or satisfies regulatory requirements.

Techno-economic evaluation provides a disciplined method for comparing alternatives. It combines an engineering model of energy use with an economic model of investment, operating cost, savings, uncertainty, and time. The purpose is not merely to calculate a payback period; it is to determine whether a project creates value under realistic technical and managerial constraints.

This chapter develops methods for estimating capital and operating costs, evaluating savings, applying payback, life-cycle cost (LCC), net present value (NPV), internal rate of return (IRR), and sensitivity analysis, and prioritizing projects when capital or operating capacity is limited.

Learning objectives

After studying this chapter, a student should be able to:

1. Distinguish capital cost, operating cost, savings, and cash flow.
2. Construct a technically defensible baseline and estimate energy and demand savings.
3. Apply simple and discounted payback, LCC, NPV, and IRR correctly.
4. Treat escalation, taxes, residual value, maintenance, degradation, and uncertainty appropriately.
5. Perform sensitivity and scenario analysis.
6. Rank and select projects subject to budget, operational, and implementation constraints.
7. Communicate assumptions, risks, and recommendations professionally.

1. The decision framework

A sound evaluation follows a sequence:

Problem definition → baseline measurement → technical options → cost and savings model → cash-flow evaluation → risk analysis → constrained selection → implementation and verification.

The baseline is the reference condition against which savings are measured. It should represent the service level that would otherwise have been delivered, not simply the most recent energy bill. For example, a chiller retrofit must be compared with the electricity that the existing chiller would have used while meeting the same cooling load and comfort requirements.

A useful baseline includes production, operating hours, weather, occupancy, set points, equipment condition, energy prices, demand charges, and maintenance practice. Where consumption depends on an independent variable, such as production, a normalized model may be used:

Here, E_b is baseline energy consumption in kWh per period, E₀ is fixed consumption, m is energy intensity in kWh per unit of activity, and Q is production or another relevant activity variable. The relationship is applicable only when the operating range is sufficiently stable and the linear approximation is supported by data.

Savings should be calculated as the difference between adjusted baseline and post-project consumption:

where S_E,t is energy saving in period t, E_b,t^adj is baseline energy adjusted for actual conditions, and E_p,t is measured post-project energy use. If demand charges apply, the monetary saving is not simply energy saving multiplied by an average tariff; energy and demand components must be separated.

2. Cost estimation for energy projects

2.1 Capital cost

Capital cost is the expenditure required to purchase and place an asset into service. It may include equipment, design, procurement, installation, civil work, controls, commissioning, training, taxes, and contingency. A practical estimate is:

The terms represent equipment, engineering, installation, civil work, control systems, commissioning, training, taxes, and contingency, respectively. All terms must be stated in the same currency and price year. Contingency should reflect identified technical and execution uncertainty; it should not be used to hide poorly defined scope.

Cost estimates have different accuracy classes. A conceptual estimate may have wide uncertainty, whereas a vendor quotation and detailed design estimate are narrower. Students and practitioners should record the estimate basis, quotation validity, excluded items, currency, escalation date, and expected accuracy range.

2.2 Operating and maintenance cost

Operating expenditure includes electricity, fuel, water, consumables, labor, service contracts, software licenses, insurance, and maintenance. The incremental annual operating cost of an option is often represented as:

A project can save energy but increase maintenance or auxiliary electricity. Therefore, annual net cash benefit should include both energy savings and non-energy effects:

Here, B_t is net operating benefit in year t, R_t is other incremental revenue or avoided cost, S_t is gross energy-related saving, and the final terms are incremental operating and tax costs. The exact tax treatment depends on the organization and jurisdiction.

2.3 Energy and demand savings

For an electrical measure, annual energy saving can be estimated from power reduction and operating time:

where S_E is kWh/year, ∆P is average real-power reduction in kW, and h is operating hours per year. This formula assumes the power reduction is approximately constant during the stated hours. If load varies, the saving should be integrated or calculated by time interval:

Power-factor correction may reduce apparent power and demand penalties, but it does not automatically reduce kWh consumption. Similarly, a motor replacement may change efficiency, loading, speed, and maintenance requirements; the saving should be based on measured or modeled operating points rather than nameplate efficiency alone.

For a tariff with energy price c_E in currency/kWh and demand price c_D in currency/kW-month, annual bill saving may be approximated by:

where ∆P_D is the reduction in billed demand in kW and S_other includes avoided fuel, water, maintenance, or other monetary effects. This expression assumes the tariff structure and demand reduction are understood; complex tariffs require interval simulation.

3. Time value of money

A currency received today is generally more valuable than the same currency received later because it can earn a return, inflation reduces purchasing power, and future cash flows are uncertain. If i is the discount rate per period, the present value of a future cash flow [[MATH{CF_t}]] is:

The discount rate must match the cash-flow convention. Annual cash flows require an annual rate; nominal cash flows containing inflation require a nominal discount rate, while real cash flows expressed in constant prices require a real discount rate. A commonly used relationship is:

where i_n is nominal discount rate, i_r is real discount rate, and f is inflation rate. Mixing nominal cash flows with a real discount rate creates a systematic error.

The discount rate may be based on the organization’s weighted average cost of capital (WACC), adjusted for project risk. WACC is not automatically appropriate for every project: a project with materially different risk may require a risk-adjusted rate or explicit scenario treatment.

4. Economic evaluation methods

4.1 Simple payback period

Simple payback is the time required for cumulative undiscounted savings to recover initial investment. For constant annual net benefit:

where PBP is in years, C₀ is initial investment, and B is constant annual net benefit. If cash flows vary, cumulative cash flow should be tabulated and interpolation used between years.

Payback is easy to communicate and useful as a screening measure. Its limitations are serious: it ignores cash flows after payback, usually ignores time value of money, and does not directly measure value creation. It is therefore unsuitable as the sole investment criterion.

4.2 Discounted payback

Discounted payback is the time at which cumulative discounted benefits equal the initial investment. Each benefit is first converted using PV_t = B_t/(1+i)^t. It recognizes time value but still ignores benefits after recovery.

4.3 Net present value

NPV is the present value of all project cash flows, including the initial investment:

Here, N is project life in years, CF_t is net cash flow in year t, RV_N is residual or salvage value at the end of year N, and i is the discount rate. If a cash flow occurs at time zero, it is included without discounting.

For mutually independent projects, a positive NPV indicates that the project is expected to earn more than the required return under the stated assumptions. For mutually exclusive alternatives, the alternative with the greatest NPV is generally preferred, provided they deliver equivalent service and risk is comparable.

For a constant annual benefit with no residual value, the present value may be calculated using the annuity factor:

This special case assumes equal end-of-year benefits, constant discount rate, and no mid-life replacement.

4.4 Internal rate of return

IRR is the discount rate that makes NPV equal to zero:

A project is normally acceptable when IRR exceeds the required rate of return, subject to risk and ranking limitations. IRR can be misleading for mutually exclusive projects of different scale, and unusual cash-flow patterns may produce multiple IRRs or no meaningful IRR. NPV should normally govern value comparison.

4.5 Life-cycle cost

LCC evaluates the total cost of owning and operating an asset over its analysis period:

C_rep,t is replacement expenditure and RV_N is residual value. LCC is particularly useful when alternatives provide the same service but have different purchase prices, efficiencies, maintenance needs, and lives. If lives differ, an equivalent annual cost or a common study period may be required.

For a present value [[MATH{PV}]] over N years, equivalent annual cost is:

This converts the present value into a uniform annual amount at discount rate i. It is applicable when comparing alternatives on an equivalent service basis.

5. Worked example: variable-speed drive retrofit

A pump operates 4,000 h/year. A variable-speed drive and control upgrade costs $48,000 installed. The measured average demand reduction is 8 kW. Electricity costs $0.12/kWh, and annual maintenance saving is $1,000. The project life is five years, the discount rate is 10%, and residual value is zero. Assume constant savings and end-of-year cash flows.

Annual energy saving is:

Annual electricity saving is:

Thus annual net benefit is:

Simple payback is:

The simple payback exceeds the five-year study life, so the project would be rejected under a five-year payback rule. However, the economic result should be checked using NPV. The five-year annuity factor at 10% is approximately 3.7908:

The NPV is negative. The project does not recover its required return under these assumptions. This example illustrates why a technically sound measure should not be approved without examining utilization, tariff structure, control performance, and capital cost. If a demand-charge saving, production benefit, or longer useful life exists, those items should be quantified rather than assumed.

6. Sensitivity, uncertainty, and risk

An economic model is not a prediction with certainty; it is a structured representation of assumptions. Important uncertain variables include capital cost, operating hours, energy price, savings rate, equipment life, discount rate, maintenance cost, implementation delay, and performance degradation.

One-way sensitivity analysis changes one variable while holding others constant. For example, NPV can be calculated for energy prices of $0.08, $0.12, and $0.16/kWh. A tornado chart can rank variables by their effect on NPV. Scenario analysis changes several coherent assumptions together, such as low, base, and high energy-price scenarios.

For a saving that decreases at annual degradation rate d, the saving in year t may be modeled as:

This is appropriate when degradation is approximately multiplicative and measurable. Reliability analysis may use expected values, but expected value should not conceal catastrophic operational consequences. A project with a high expected NPV but a significant probability of production interruption may require mitigation, insurance, redundancy, or rejection.

Decision makers should distinguish:

• Risk: outcomes have identifiable probabilities or distributions.
• Uncertainty: probabilities are poorly known.
• Sensitivity: response of the result to a changed assumption.
• Scenario: a coherent set of assumptions about future conditions.

Monte Carlo simulation can assign distributions to uncertain inputs and produce a distribution of NPV rather than a single number. The quality of the result depends on input data, correlations, and model structure; simulation is not a substitute for engineering judgment.

7. Project prioritization under constraints

Organizations rarely possess unlimited capital, skilled labor, shutdown time, or plant space. Project selection is therefore a constrained optimization problem rather than a simple ranking exercise.

A first screen may exclude projects that fail safety, legal, reliability, or minimum technical requirements. Among feasible projects, NPV is a strong value measure. For a single capital constraint, a useful screening index is profitability index:

A PI greater than one indicates positive NPV when the denominator is the initial investment and there are no special constraints. However, ranking by PI alone can be suboptimal because it may select many small projects while excluding a larger combination with greater total NPV.

A binary selection model can express the decision:

subject to:

and, for a labor constraint:

where x_j is 1 if project j is selected and 0 otherwise, B is available capital, L_j is labor requirement, and L_max is available labor. Additional constraints may represent shutdown windows, site capacity, dependencies, or mutually exclusive alternatives. For a small number of projects, enumeration or spreadsheet optimization is adequate; larger problems may use integer programming.

Operational constraints must be treated as real constraints. A project that requires a shutdown unavailable during peak production is not feasible merely because its NPV is high. A technically inferior project may become preferable if it can be installed during routine maintenance and carries less implementation risk.

8. Illustrative case: industrial facility portfolio

This is a fictional illustrative case. An industrial facility has $100,000 available for energy projects during an annual shutdown. Three independent proposals are prepared:

Assume end-of-year benefits, no residual value, and a total labor capacity of 300 h. Project combinations within the capital limit are A+B ($65,000; 200 h; NPV $22,198) and C alone ($75,000; 240 h; NPV $8,477). A+C exceeds the capital limit, while B+C costs $100,000 and requires 320 h, exceeding labor capacity. All three exceed capital.

The financially best feasible combination is A+B, not the single project with the largest annual saving and not necessarily the project with the highest individual NPV. The recommendation should still be conditional on confirming the baseline, vendor scope, savings persistence, and shutdown schedule. Measurement and verification should be planned before implementation so that realized savings can be compared with the approved case.

The case demonstrates four professional practices: evaluate portfolios rather than isolated projects, include operational constraints, use NPV for value comparison, and preserve an implementation contingency for uncertain scope.

9. Measurement, verification, and governance

Economic approval should identify how savings will be verified. Common practice compares measured post-project performance with an adjusted baseline, accounting for production, weather, occupancy, and operating hours. Metering boundaries, data quality, calibration, reporting frequency, and responsibility should be documented.

Governance also requires transparent assumptions. Analysts should disclose whether savings are gross or net, whether taxes and incentives are included, how avoided maintenance is supported, and whether costs are nominal or real. Incentives should be treated consistently: an upfront grant reduces initial investment, whereas a taxable incentive may affect after-tax cash flow differently.

Ethically, an energy manager must not inflate savings, suppress unfavorable sensitivity results, or select a discount rate merely to force approval. Procurement should avoid conflicts of interest, preserve competitive fairness, and distinguish vendor claims from independently verified performance. Safety, worker welfare, cybersecurity of connected controls, and compliance with electrical and environmental regulations are part of the decision, even when they are not easily represented by NPV.

Artificial intelligence can assist with load forecasting, anomaly detection, tariff classification, and probabilistic savings estimation. It should be used as decision support, not as an authority. Training-data bias, unexplained recommendations, cybersecurity, privacy, and model drift require human review. An AI-generated saving estimate must be validated against meters, engineering constraints, and accountable assumptions.

10. Best-practice checklist

1. Define the service requirement and decision boundary.
2. Establish a measured, normalized baseline.
3. Separate energy, demand, maintenance, production, and environmental effects.
4. Estimate total installed cost, not only equipment price.
5. State price year, currency, tariff, tax, incentive, and escalation assumptions.
6. Match nominal or real cash flows with the corresponding discount rate.
7. Use NPV as the principal value criterion; use payback as a communication or screening metric.
8. Test capital cost, savings, operating hours, tariff, life, and delay sensitivity.
9. Check reliability, safety, cybersecurity, and shutdown feasibility.
10. Select portfolios under actual capital, labor, and operational constraints.
11. Document risks, responsibilities, and approval gates.
12. Verify post-installation performance and update the economic model.

Challenges and limitations

No financial method eliminates poor data. Baseline errors, changing production, tariff reforms, rebound effects, equipment interactions, and uncertain maintenance can dominate calculated returns. Simple payback can favor short-lived measures, while IRR can mis-rank projects of different scale. NPV depends on the discount rate and analysis period, and LCC depends on service equivalence and reliable future cost assumptions. Non-monetary benefits such as resilience, worker comfort, emissions reduction, and regulatory compliance should be described explicitly and, where defensible, valued through a multi-criteria decision process rather than ignored.

Chapter Summary

Techno-economic evaluation links engineering performance to financial decision-making. A defensible analysis begins with a measured and adjusted baseline, estimates total installed cost and incremental operating cost, and distinguishes energy savings from demand and non-energy benefits. Simple and discounted payback are accessible screening tools, but they ignore some or all post-recovery value. LCC compares the present cost of alternatives delivering equivalent service, while NPV provides the strongest general measure of value creation. IRR is useful as a rate-of-return indicator but has important ranking and cash-flow limitations.

Discount rates, inflation, escalation, residual value, degradation, taxes, and replacement costs must be treated consistently. Sensitivity and scenario analysis expose which assumptions control the decision. Finally, project selection must account for capital, labor, shutdown, reliability, safety, and regulatory constraints. Portfolio optimization, measurement and verification, ethical reporting, and responsible use of digital and AI tools convert an economic calculation into a professional project decision.

Key Terms

Baseline: Reference energy or operating condition used to calculate savings.

Capital cost: Initial expenditure required to purchase and commission a project.

Operating cost: Recurring cost of operating, maintaining, and supporting an asset.

Cash flow: Time-stamped inflow or outflow of money associated with a project.

Simple payback: Initial investment divided by undiscounted annual net benefit for constant benefits.

Discounted payback: Time required for discounted benefits to recover initial investment.

Life-cycle cost: Present cost of acquisition, operation, maintenance, replacement, and disposal less residual value.

Net present value: Present value of all project cash flows, including initial investment.

Internal rate of return: Discount rate that makes project NPV equal to zero.

Discount rate: Required return used to convert future cash flows to present value.

Nominal cash flow: Cash flow expressed in money of the year in which it occurs, including inflation effects.

Real cash flow: Cash flow expressed in constant-price purchasing power.

Residual value: Value remaining at the end of the analysis period.

Sensitivity analysis: Study of result changes caused by varying one or more assumptions.

Scenario analysis: Evaluation of coherent alternative future conditions.

Profitability index: Present value of inflows divided by initial investment.

Measurement and verification: Process for determining whether predicted savings were achieved.

Constraint: Limitation on capital, labor, time, space, or operations that affects project selection.

Review Questions

Short Answer Questions

1. Why is a technically feasible energy project not necessarily an economically acceptable project?
2. Define a baseline and explain why it must be adjusted for operating conditions.
3. State two limitations of simple payback.
4. What is the difference between nominal and real cash flows?
5. Why should demand savings be separated from energy savings in tariff analysis?

Long Answer Questions

1. Explain the complete process for evaluating an industrial energy-efficiency project from baseline development through post-installation verification.
2. Compare payback, LCC, NPV, and IRR, including the strengths and limitations of each method.
3. Explain how inflation, discount rate, escalation, residual value, and degradation affect a life-cycle economic model.
4. Discuss how technical, operational, safety, ethical, and regulatory considerations influence project approval.
5. Describe how a portfolio of energy projects can be selected when capital and shutdown labor are limited.

Analytical / Application Questions

1. A motor retrofit costs $30,000 and saves 6 kW for 3,500 h/year at $0.10/kWh. It also reduces maintenance by $500/year. Calculate annual benefit and simple payback, then identify two additional data items needed before approval.
2. Two mutually exclusive chiller alternatives have different lives and maintenance costs. Develop an LCC or equivalent-annual-cost comparison and state the assumptions required for a fair decision.
3. An approved project has negative NPV when energy prices are low but positive NPV when prices are high. Design a sensitivity and risk-mitigation plan, including at least four variables and a measurement-and-verification strategy.

Multiple Choice Questions with Answers

Question 1

Which measure is generally the strongest indicator of value creation for an independent energy project?

A. Simple payback period

B. Net present value

C. Equipment purchase price

D. Annual operating hours

Correct Answer: B. Net present value

Question 2

A project reduces average demand by 5 kW for 2,000 h/year. Its annual energy saving is:

A. 400 kWh

B. 2,000 kWh

C. 10,000 kWh

D. 40,000 kWh

Correct Answer: C. 10,000 kWh

Question 3

If an investment is $20,000 and constant annual net benefit is $5,000, the simple payback is:

A. 2 years

B. 4 years

C. 5 years

D. 25 years

Correct Answer: B. 4 years

Question 4

In an NPV calculation, a positive NPV normally means that the project:

A. Has no operating cost

B. Recovers its investment and exceeds the required return

C. Has the shortest payback

D. Has zero technical risk

Correct Answer: B. Recovers its investment and exceeds the required return

Question 5

Which item belongs in total installed capital cost?

A. Only the equipment catalogue price

B. Only annual electricity savings

C. Engineering, installation, commissioning, and appropriate contingency

D. Only the project’s residual value

Correct Answer: C. Engineering, installation, commissioning, and appropriate contingency

Question 6

The IRR is defined as the discount rate that:

A. Maximizes annual energy saving

B. Makes NPV equal to zero

C. Equals the electricity tariff

D. Minimizes capital cost

Correct Answer: B. Makes NPV equal to zero

Question 7

Which statement about simple payback is correct?

A. It always accounts for cash flows after recovery

B. It automatically includes inflation

C. It is easy to use but generally ignores time value of money

D. It is identical to NPV

Correct Answer: C. It is easy to use but generally ignores time value of money

Question 8

A baseline should be adjusted primarily to:

A. Make savings appear larger

B. Reflect changes in relevant operating conditions

C. Remove all maintenance costs

D. Replace measured data with vendor claims

Correct Answer: B. Reflect changes in relevant operating conditions

Question 9

For a project with uncertain energy prices and operating hours, the most direct first analysis is:

A. Sensitivity or scenario analysis

B. Ignoring uncertainty

C. Using equipment nameplate power only

D. Reporting only the shortest payback

Correct Answer: A. Sensitivity or scenario analysis

Question 10

When selecting projects under a capital and labor limit, the preferred approach is to:

A. Select the project with the largest equipment size

B. Rank only by annual energy saving

C. Evaluate feasible project combinations using value and constraints

D. Select every project with a positive simple payback

Correct Answer: C. Evaluate feasible project combinations using value and constraints