How AEC Savings Are Calculated: Baselines and New Construction

Every Alternative Energy Credit in Pennsylvania traces back to one subtraction: what the building would have used, minus what it actually uses. The second number is measured. The first is chosen — and the rules that govern that choice are where most of the disagreement in an AEC application lives.

The Pennsylvania Technical Reference Manual calls that chosen number the baseline, and it does not use one. It uses four, assigned by what the manual calls measure vintage. Getting the vintage right is not a paperwork detail. It determines the size of the savings number, whether that number holds steady or steps down partway through the measure's life, and in the case of a new building, whether anyone thought to establish a baseline before the walls went up.

The four measure vintages

TRM Volume 1, Section 1.4 defines them. Every project falls into one.

• Retrofit (RET) — "Measures that modify or add on to existing equipment with technology to make the system more energy efficient."
• Early Replacement (EREP) — "The replacement of existing equipment, which is functioning as intended and is not operationally obsolete, with a more efficient model primarily for purposes of increased efficiency."
• Replace on Burnout (ROB) — replacement at the point of natural failure.
• New Construction (NC) — "The substitution of efficient equipment for standard baseline equipment that the customer does not yet own or during the course of a major renovation project that removes existing, but operationally functional equipment."

Volume 1, Section 1.8 groups these into two economic families, and the grouping explains the logic:

"Most measures will fall into one of two categories, each with a baseline that is most commonly used: One for market-driven choices — often called 'lost opportunity' and either replacing equipment that has failed (replace on burnout) or new installations (new construction). One for discretionary installations — either early replacement or retrofit."

The distinction is about counterfactuals. When equipment fails, or when a building is being built, the owner is buying *something* regardless. The only question is how efficient that something is, so the baseline is what code or common practice would have delivered. When an owner rips out working equipment early, they are making a discretionary choice, and the counterfactual is that the old equipment kept running.

Retrofit and early replacement: the dual baseline

This is the mechanism most often missing from project models, and it cuts savings in half partway through.

The TRM's definition of a retrofit measure continues: "Retrofit measures may have a dual baseline. For the estimated remaining useful life of the existing equipment the baseline is the existing equipment; afterwards, the baseline is the applicable code, standard, or industry standard practice expected to be in place at the time the unit would have been naturally replaced or retrofitted." The early replacement definition carries identical language.

Volume 1, Section 1.16 restates it: "avoided cost savings for programs where measures replace units before the end of their useful life are measured from the efficient unit versus the replaced unit for the remaining useful life of the existing unit, then from the efficient unit versus a new standard unit for the balance of the efficient measure's life."

In practice this means a retrofit's savings curve has a step in it. For the first stretch — the remaining useful life of whatever was torn out — savings are measured against genuinely old, inefficient equipment, and they are large. After that point the baseline resets to code, and the savings for the remainder of the measure's life are only the margin between the installed equipment and a code-minimum unit. The credit stream drops without anything physical changing.

The manual acknowledges this is awkward to administer. Both definitions add: "In practice, in order to avoid the uncertainty surrounding the determination of 'remaining useful life,' measure savings and costs sometimes follow ROB baseline and incremental cost definitions." In other words, some programs skip the dual baseline and use the code baseline throughout — which is more conservative and simpler, and produces a smaller first-years number.

If a project's AEC model shows a flat annual credit figure for fifteen years off a retrofit, ask which baseline treatment it assumed. The answer changes the total.

New construction: code from day one, and no step down

A new building has no existing equipment, so there is nothing to measure against. The TRM resolves this by naming the counterfactual directly.

From the New Construction definition: "The baseline used for calculating energy savings is the construction of a new building or installation of new equipment that complies with applicable code, standard or industry standard practice in the absence of applicable code or standards in place at the time of construction/installation/substantial renovation."

Section 1.8 adds the Commission's preference: "For all new construction (NC) and ROB scenarios, the baseline is typically a jurisdictional code or a national standard; however, there may be cases where a market baseline is appropriate. In these scenarios, the Commission prefers codes and standards as it is too expensive and time consuming to conduct market baseline and characterization research."

Which code, specifically. Volume 1, Section 1.8 states that Pennsylvania has been adopting the 2021 International Energy Conservation Code, with a scheduled effective date of 13 July 2025, and that "commercial protocols relying on code standards as the baseline condition may refer to either IECC 2021 or ASHRAE 90.1-2019 per the program design."

Here is the part that matters and gets missed: the New Construction and Replace on Burnout definitions contain no dual-baseline language. The code baseline holds for the full measure life. New construction savings are narrower than a retrofit's opening years — code has already captured most of the easy efficiency — but they do not step down. A retrofit's credit stream shrinks mid-life. A new building's does not.

For a project with a long measure life, that flat profile can be worth more in aggregate than a front-loaded retrofit curve, even though the first-year number looks less impressive. Compare the two on lifetime volume, never on year one. Our breakdown of measure life by project type gives the multiplier for that calculation.

The permit date locks your baseline

This is a genuine timing lever and almost nobody uses it deliberately.

TRM Volume 1 states: "For new construction, the appropriate TRM must be based on the date when the building/construction permit was issued (or the date construction starts if no permit is required) because that aligns with codes and standards that define the baseline. Savings begin to accrue at the project's ISD."

Codes tighten over time. A tighter code is a more efficient baseline, which means a smaller measured gap between baseline and as-built, which means fewer credits for the same building. A project permitted before a code change is measured against the older, looser baseline for its entire measure life.

So the permit date is not merely an administrative milestone. It fixes the reference point that every credit for the next fifteen years is calculated against. On a project already in design near a code transition, that date is worth understanding before it is set.

How new construction savings are actually computed

Volume 3, Section 3.1.2 covers New Construction Lighting — the clearest worked example in the manual, carrying a 15-year measure life.

The measure exists, in the TRM's words, "to encourage decision-makers in new construction projects to incorporate greater energy efficiency into their building design and construction practices that will result in a permanent reduction in electrical (kWh) usage above baseline practices."

Eligibility is broad: "New construction applies to new building projects wherein no structure or site footprint presently exists, addition or expansion of an existing building or site footprint, or major tenant improvements that change the use of the space." Additions and change-of-use tenant fit-outs count, which surprises people.

The baseline is constructed rather than observed: "The baseline demand for calculating savings is determined using one of the two methods detailed in IECC 2021. The interior lighting baseline is calculated using either the Building Area Method... or the Space-by-Space Method. For exterior lighting, the baseline is calculated using the Baseline Exterior Lighting Power Densities."

Under the Building Area Method, the baseline connected load is the building area multiplied by a Lighting Power Density for the building type. A sample of the IECC 2021 values used, in watts per square foot: office 0.64, retail 0.84, warehouse 0.45, hospital 0.96, manufacturing facility 0.82, school or university 0.72, parking garage 0.18.

Savings are then the difference between that constructed baseline load and the as-built load, carried across operating hours and adjusted for savings from controls and for interactive effects with the HVAC system.

The practical consequence: on a new building, the savings number is decided at design, not at commissioning. A warehouse designed at 0.30 W/ft² against a 0.45 baseline has locked in its credit volume before a fixture is hung. There is no later opportunity to improve the number, because the baseline is fixed by code and the design is fixed by the drawings.

Measurement: where defaults stop being allowed

TRM Volume 1, Table 1-2 sets the kWh savings thresholds at which measurement becomes mandatory. C&I solar PV at 2,000,000 kWh; C&I lighting at 750,000; C&I HVAC, motors and VFDs, building shell and agricultural equipment each at 250,000.

For lighting above the threshold, Volume 3 requires metering, and then writes an exception specifically shaped for new buildings: "Exceptions may be made, and EDC data gathering may be substituted, if necessary, at the evaluation contractor's discretion in cases involving early occupancy. Otherwise, installation of light loggers is the accepted method of metering, but trend data from BMS is an acceptable substitute."

That BMS allowance is the practical route for a new building, whose building management system is commissioned and trending before the first tenant moves in. The manual attaches conditions: "The BMS trends should represent the actual status of the lights (not just the command sent to the lights)," and the data "should be specific to the lighting systems and should be required to be representative of the building areas included in the lighting project."

A new building that specifies its BMS points with the AEC application in mind satisfies this at essentially no incremental cost. One that does not is retrofitting loggers into a finished building. See submetering across multi-measure projects for how to isolate savings when several measures land at once.

For projects outside the deemed and partially deemed protocols, Volume 1, Section 1.17 directs that "site-specific custom measure protocols be established in general conformity to the International Performance Measurement and Verification Protocol (IPMVP) or Federal Energy Management Program M&V Guidelines," and states plainly that "the qualification for and availability of AEPS Credits and energy efficiency and demand response savings are determined on a case-by-case basis."

One honest discrepancy worth naming

We are flagging this rather than smoothing it, because a reader will hit it.

The AEPS Program Administrator's public-facing description of energy efficiency states that credits "are generated based on the savings between an inefficient technology and the new efficient technology," and gives the formula as pre-retrofit kWh less post-installation kWh. That formula is undefined for a new building. There is no pre-retrofit consumption.

Against that, the Administrator's own Energy Efficiency Information Guide — the sheet listing required application attributes — asks the applicant to select a Measure Vintage from "Replace on Burnout (ROB), New Construction, Retrofit (RET), or Early Replacement Measure (EREP)," and a Program from a list that includes "Residential New Construction" and "Commercial and Industrial Construction." New Construction is an explicitly selectable vintage on the application itself, and the TRM that the guide points to defines its baseline.

The application form is the operative document and the marketing page is a simplification. But we found no statutory or regulatory sentence in Act 213 or 52 Pa. Code Chapter 75 that grants or denies new construction eligibility — the Act is silent on measure vintage entirely. The support for new construction is administrative and structural, not statutory. Anyone modelling a new building's AEC revenue should confirm the vintage with the Administrator in writing before the numbers reach a pro forma. Our overview of Act 213 and the legal foundation of the Tier II market covers what the statute does and does not say.

What this changes about project selection

• On a retrofit, ask about the dual baseline. A flat fifteen-year credit projection off a retrofit is either assuming the ROB simplification or ignoring the step down. Both are defensible; neither should be silent.
• On a new building, the AEC conversation belongs in design review. Baseline is code, savings are the design margin, and the margin is fixed once the drawings are. There is no post-hoc improvement path.
• Watch the permit date near a code change. It fixes the baseline for the full measure life.
• Specify the metering before you need it. The thresholds are lower than most people assume — 250,000 kWh for HVAC and motors — and BMS trend data is accepted where it is properly scoped.

For which project types qualify in the first place, start with qualifying projects for Tier II credits, and see how enrollment works for the sequence.

Frequently Asked Questions

What is a baseline in an AEC savings calculation?

The baseline is the energy consumption a project is measured against — what the building would have used absent the efficiency measure. The Pennsylvania TRM assigns a baseline according to measure vintage: existing equipment for retrofits and early replacements, and applicable code or industry standard practice for new construction and replace-on-burnout projects.

Does new construction qualify for PA Tier II AECs?

The AEPS Program Administrator's Energy Efficiency Information Guide lists New Construction as a selectable measure vintage and includes Residential New Construction and Commercial and Industrial Construction among its programs, and the TRM defines a New Construction baseline. However, neither Act 213 nor 52 Pa. Code Chapter 75 addresses measure vintage, so the support is administrative rather than statutory. Confirm eligibility with the Administrator in writing before modelling revenue.

What is the baseline for a new construction project?

Code. The TRM defines it as construction or equipment that complies with the applicable code, standard, or industry standard practice in place at the time of construction. For commercial buildings that means IECC 2021 or ASHRAE 90.1-2019, and for lighting the baseline load is calculated from Lighting Power Densities for the building type.

What is a dual baseline?

For retrofit and early replacement measures, the TRM applies the existing equipment as the baseline for the estimated remaining useful life of that equipment, then switches to the applicable code or standard for the balance of the new measure's life. Savings are larger in the first period and smaller after the switch. New construction and replace-on-burnout measures do not use a dual baseline.

Does the permit date affect how many AECs a new building earns?

Yes. The TRM states that for new construction the governing manual is set by the date the building or construction permit was issued, because that aligns with the codes and standards defining the baseline. A project permitted before a code tightening is measured against the looser baseline for its entire measure life.

When is metering required rather than optional?

TRM Volume 1, Table 1-2 sets savings thresholds: 2,000,000 kWh for commercial and industrial solar PV, 750,000 kWh for lighting, and 250,000 kWh each for HVAC, motors and VFDs, building shell, and agricultural equipment. Above the threshold, measurement is required. For lighting, building management system trend data is an accepted substitute for light loggers.

*Sources: 2026 Pennsylvania Technical Reference Manual, Volumes 1 and 3, Pennsylvania Public Utility Commission, adopted 12 September 2024; PennAEPS Energy Efficiency Information Guide; 52 Pa. Code Chapter 75. Nothing in this article is legal, tax or investment advice.*