ASHRAE Level 2 Energy Audit: What Owners Get for the Money

An owner asks three firms to quote an energy audit on the same 180,000 square foot office building and gets back three numbers that are not within a factor of five of each other. The instinct is to assume the cheap one is efficient and the expensive one is padded. Usually neither is true. The three firms have quoted three different audit levels, all legitimate, and none of the proposals said so clearly enough for the owner to notice.

The levels are not marketing tiers. They are defined in a published standard, they have specific required content, and each exists to support a different decision. Buying the wrong one is expensive in both directions: a walk through report cannot justify a capital request, and a full investment grade study on a building nobody has screened is money spent proving something a tenth of the fee would have shown. Budlong delivers MEP engineering services and existing building energy work across California, and what follows comes from reviewing a great many audit reports written by other people. You can see the range of that work on our expertise overview.

1. What Is an ASHRAE Level 2 Energy Audit?

A Level 2 energy audit is an energy survey and engineering analysis of an existing building, carried out to the content requirements of ANSI/ASHRAE/ACCA Standard 211, the Standard for Commercial Building Energy Audits. It lists specific energy conservation measures, each with annual energy saved, annual cost saved, cost to implement and simple payback, plus the analysis behind those numbers.

What matters is the decision it is fit to support. A Level 2 will carry a capital plan, a utility incentive application and a five year budget. It will not, on its own, carry a guaranteed savings contract on a multi million dollar plant replacement.

Why the standard exists

Before Standard 211, the level names came from an earlier ASHRAE procedures document that described the tiers without compelling any content, so two firms could both sell a Level 2 and deliver very different work. Standard 211 turned the descriptions into requirements with defined minimum content and standardised reporting forms, which is what makes proposals comparable. It is maintained through the ASHRAE standards and guidelines process. A proposal that says “Level 2 audit” without citing Standard 211 is describing a price point, not a scope.

2. How the Level 2 Analysis Actually Works

A Level 2 is a sequence, and each stage constrains the next. Knowing it is how you tell a real audit from a walk through with a spreadsheet attached.

Utility bill analysis and normalisation

Standard 211 requires at least twelve consecutive months of billing data, and up to three years is usual, for every meter and fuel. The analysis separates consumption from demand, identifies the rate schedule and time of use periods that decide what a saved kilowatt hour is worth, normalises for weather using degree days, and calculates energy use intensity. A building whose load is flat all year is telling you something different from one whose summer peak triples.

Load disaggregation and the end use breakdown

This is the analytical core of a Level 2 and the stage most often skipped. Consumption is broken into end uses: heating, cooling, fans and pumps, lighting, hot water, refrigeration, plug and process loads. Standard 211 requires that breakdown to account for between 90 and 100 percent of historical billed energy.

The reconciliation is the point. If the end use model says 4.2 million kilowatt hours and the meter says 5.6 million, something is running that nobody has accounted for, and finding it is often worth more than every measure in the report combined.

The site survey and systems inventory

The survey is a full inspection, not a tour. It records major equipment with nameplate capacity, efficiency, installation date and condition; schedules as actually run rather than as written; and control sequences as found, including every override still in place. Standard 211 expects key operating parameters for the systems covering more than 80 percent of energy use.

Spot measurement and short term metering

Measurement is the main thing separating Level 2 from Level 1: clamp on power readings at panels and motors, temperature and humidity logging, carbon dioxide logging where demand control ventilation is in play, light levels, combustion analysis, and trend logs from the building automation system over days or weeks. This is where the discretionary budget sits, so a proposal that states how many points will be logged, and for how long, is a better proposal than one that stays silent.

From findings to measures

The last stage converts observations into measures. Each gets an engineering calculation of annual energy saved by fuel, a cost saving computed against the actual rate schedule rather than a blended average, an implementation cost, a simple payback, and a life cycle metric such as savings to investment ratio. Interacting measures should be adjusted, not summed.

3. The Four Tiers of the ASHRAE Audit Framework

There are four tiers, not three, and the first is nearly free.

Preliminary energy use analysis and benchmarking

Twelve months of consumption data, up to three years, produce an energy use intensity compared against peers of similar type, size and climate, normally through ENERGY STAR Portfolio Manager, whose peer data comes from the Commercial Buildings Energy Consumption Survey. It does not tell you what is wrong. It tells you whether it is worth finding out.

Level 1: walk through analysis

A brief site visit, staff interviews, a closer read of the bills, and a report listing low cost and no cost opportunities plus capital projects worth further study. Savings and costs are qualitative or broad ranges. It answers whether there is enough here to justify a deeper look.

Level 2: energy survey and engineering analysis

All Level 1 content plus the end use breakdown, the detailed survey, measurement and quantified measure economics. It is the level most owners should buy most of the time.

Level 3: detailed analysis of capital intensive modifications

All Level 2 content, then depth on a small number of large measures: extended metering over weeks or months, calibrated simulation, contractor pricing rather than estimates, and a risk assessment. It is often called an investment grade audit because a lender or a savings guarantee can be written against its numbers.

4. Which Buildings and Decisions Trigger a Level 2

The California regulatory layer

California does not require energy audits statewide. It requires benchmarking and disclosure. Under Assembly Bill 802 the California Energy Commission requires annual benchmarking through ENERGY STAR Portfolio Manager, submitted by 1 June each year, for nonresidential buildings over 50,000 square feet with no residential utility accounts and multifamily buildings over 50,000 square feet with seventeen or more residential accounts. Results are published on a public dashboard, so a poor energy use intensity is visible to tenants and lenders.

Audit mandates come from local ordinances instead. The largest is the Los Angeles Existing Buildings Energy and Water Efficiency programme, covering buildings above 20,000 square feet, requiring annual benchmarking by 1 June and an energy and water efficiency audit or retro-commissioning on a five year cycle, with exemptions for buildings that can demonstrate strong performance. Other jurisdictions run their own programmes, so this is a local question rather than a state one.

The commercial triggers

Most Level 2 audits are not driven by an ordinance at all. They are driven by one of these:

• A capital plan needing defensible numbers, usually when major equipment nears end of useful life and the choice is replace in kind, upgrade, or change systems.
• Acquisition due diligence, where the buyer wants the real operating cost and the deferred maintenance behind it.
• A utility incentive application requiring documented pre-existing conditions and a savings calculation.
• A decarbonisation plan, where the question is which loads are gas and when those assets reach end of life.
• A tenant or investor performance commitment that somebody now has to produce a route to.

When a Level 2 is the wrong purchase

If the question is narrow, buy something narrower. A building whose only real question is whether to replace the chillers does not need a whole building Level 2, and if nobody has benchmarked the building at all, a Level 2 is premature.

5. What a Level 2 Asks of Each Building System

The survey is organised by system, and each has characteristic failures a competent auditor hunts for.

HVAC and central plant

Chiller and boiler efficiency at part load rather than nameplate, cooling tower approach and fan control, motors sized against actual duty, whether variable speed drives modulate or sit pinned at full speed, economiser operation, and simultaneous heating and cooling. The last of those is among the most common and most expensive findings in commercial buildings, and it rarely appears as an equipment fault.

Controls and sequences of operation

Schedules against actual use, optimal start and stop, supply air temperature and static pressure reset, chilled and hot water reset, minimum outdoor air position, demand control ventilation function, sensor calibration, and the overrides an operator added three years ago and never removed. The ASHRAE Guideline 36 sequences are the reference point most auditors measure existing logic against.

Lighting

Lamp and driver type, installed lighting power density by space type, control presence and function, daylight response, and exterior hours. Easy to quantify, and in a building already on LED, often the smallest remaining opportunity.

Domestic hot water and plumbing

Water heater efficiency and standby losses, storage temperature, whether the recirculation pump runs continuously, pipe insulation, mixing valve function, and fixture flow rates.

Electrical distribution and plug loads

Transformer loading and losses, power factor and reactive charges, the peak demand profile and whether staging could spread it, receptacle controls, and the plug and process load that in modern offices frequently exceeds lighting.

Envelope

Glazing type and condition, observable insulation, air leakage, roof condition and reflectance, and infrared survey where scope allows. Envelope measures rarely pay back on energy alone, but they change the plant sizing question when equipment is due for replacement, which is where their value sits.

6. Adjacent Standards, M&V and the California Layer

Measurement and verification, and IPMVP

An audit estimates savings. Measurement and verification proves them. The framework is the International Performance Measurement and Verification Protocol, published by the Efficiency Valuation Organization. Option A isolates the retrofit and measures the key parameter while stipulating the rest, Option B isolates it and measures all parameters, Option C uses whole facility utility data and regression, and Option D uses calibrated simulation where no reliable baseline exists.

The choice matters commercially. A control sequence measure verified under Option C vanishes into the noise of a whole building bill unless the saving is large, while a lighting retrofit under Option A is straightforward. If any part of the project carries a guarantee or a performance linked incentive, pick the option before implementation, because the baseline data has to be captured first.

Utility incentive and rebate programmes

Deemed incentives pay a fixed amount per unit for standard measures and need little analysis. Custom incentives pay against calculated savings and require exactly what a Level 2 produces: documented pre-existing conditions, a transparent calculation and often an M and V plan. The most common way owners lose an incentive is sequencing. Once equipment has been replaced the pre-retrofit baseline cannot be documented, and the application fails on evidence rather than merit.

Title 24 and the alteration trigger

The moment a recommendation becomes a permitted alteration, the California Energy Code applies to it. Replacing a rooftop unit pulls in requirements for that unit and often for the economiser and controls serving it. Replacing a defined proportion of luminaires triggers the lighting control requirements for that space. Reconfiguring ductwork triggers sealing and leakage testing. The triggers are cumulative, so a phased programme across several permits can quietly bring a whole floor into scope.

Decarbonisation planning

For an owner with a carbon target, the audit is the baseline document. The end use breakdown identifies which loads are gas, the equipment inventory identifies when each gas fired asset reaches end of life, and the electrical survey identifies whether the service can absorb the added load. That last constraint stops more electrification projects than anything else, and the combination is the practical starting point for sustainable engineering solutions on an existing asset.

7. Healthcare, Laboratories, Data Centres and Schools

Healthcare

Air change rates, pressure relationships and filtration in clinical spaces exist for infection control and are not available as energy measures. The savings sit in the plant, in heat recovery, in sequences serving non clinical areas, and in scheduling for spaces that genuinely are unoccupied. An auditor who proposes cutting air changes in an operating room has told you they do not work in healthcare facilities.

Laboratories

Laboratory energy use intensity is often several times that of an equivalent office, and the dominant load is ventilation. What matters is fume hood sash management, variable air volume hood control, occupancy based air change reduction where the safety case supports it, and exhaust heat recovery. Because the loads are so large, a Level 2 in a laboratory or industrial facility justifies more metering than an equivalent area of office.

Data centres and mission critical

The metric is power usage effectiveness, and the constraint is that nothing may be tested in a way that risks the load. Airflow management, containment, supply temperature setpoints against current equipment tolerances, economiser hours, and uninterruptible power supply efficiency at actual load factor are the recurring findings. Redundancy often leaves equipment running well below efficient loading, which is a design question rather than an operating one. Our mission critical work sits here.

Education

Schools have the clearest occupancy schedules of any building type and the largest gap between scheduled and actual hours. Summer and holiday operation, gymnasium and auditorium scheduling, portable classrooms and campus plant staging are the usual findings. Because schools and universities run many similar buildings, a sample audit extrapolates across the estate in a way that rarely works elsewhere.

8. How to Choose the Right Audit Level

The choice is about the decision the report has to survive, not about building size.

The failure mode we see most often is a Level 1 bought when a Level 2 was needed. The owner takes twenty opportunities with no numbers attached to a capital committee and is asked what each costs and saves. The report cannot say, because it was never scoped to, so the audit is redone and the second fee is not discounted for the first.

9. Reading the Report Before You Accept It

A Level 2 report is a deliverable you paid for and are entitled to reject. Most owners never review one critically, which is why so many sit unread.

What the report must contain

• A utility summary covering twelve to thirty six months, by meter and by fuel, with rate schedules identified.
• A benchmarking result with energy use intensity and the peer comparison behind it.
• An energy end use breakdown that reconciles to billed energy, with the reconciliation shown.
• An equipment inventory with nameplate data, installation dates and observed condition.
• Operating schedules and control sequences as found, not as designed.
• The measurement data actually collected, with logging period and points stated.
• A measure table giving annual energy savings by fuel, annual cost savings, implementation cost, simple payback and a life cycle metric for every measure.
• A statement of assumptions and an explicit note of where measures interact.
• A discussion of operations and maintenance improvements kept separate from the capital measures.

How to read the savings estimates sceptically

Are the savings interactive or additive? If ten measures were each calculated against the original baseline and then added, the total is overstated, sometimes substantially. Reducing lighting load reduces cooling load, so a lighting measure and a chiller measure both claim part of the same kilowatt hour. A competent report states the bundle saving separately from the sum of the parts.

What is the baseline? A baseline built from design intent, nameplate efficiency and scheduled hours produces far larger savings than one built from how the building actually runs, because the building is already failing to achieve design intent. Savings measured against a fiction are a fiction. The baseline should be the as found condition, documented.

Is this an engineering calculation or a model? Bin methods and spreadsheet analysis are appropriate at Level 2 for most measures and have the advantage of being transparent. Calibrated simulation is more accurate for measures that interact with the whole building, and it is a Level 3 activity. What is not acceptable is modelled precision presented on top of calculated inputs. Savings quoted to four significant figures are a warning, not rigour.

Low cost measures, capital measures and the operations discussion

Standard 211 expects the measure list to separate low cost or no cost items from capital items, and that split is the most commercially useful thing in the report. No cost items can usually be executed from the operating budget within weeks; capital items go into a multi year plan. Merging the two into one list sorted by payback buries what you could act on tomorrow.

Retro-commissioning is usually the highest return finding

The shortest payback measures in a Level 2 report are almost always the ones that put existing systems back into correct operation rather than replacing them. Research by Lawrence Berkeley National Laboratory across roughly 1,500 North American buildings found median existing building commissioning costs of about twenty six cents per square foot, a median simple payback of 1.7 years, and a twenty fifth to seventy fifth percentile range of 0.8 to 3.5 years. The same work put new construction commissioning near a dollar per square foot with a median payback around 4.2 years.

That is why the retro-commissioning findings deserve to be read before the capital measures. The distinction between the two processes is set out in commissioning versus retro-commissioning. Executing the findings is a commissioning services scope rather than a design scope, and it follows the building commissioning process applied to an existing asset.

The biggest savings are controls, not equipment

Owners commission audits expecting a shopping list of equipment. A properly done audit usually returns a list of sequences. Air handlers running twenty four hours in a building occupied twelve. Supply air temperature fixed at a setpoint chosen in 2009. Static pressure reset disabled after a noise complaint. Economisers locked closed since an actuator failed. Minimum outdoor air set for a worst case occupancy that happens four days a year.

None of that requires capital, all of it requires somebody competent at the building automation system, and most of it drifts back within two years unless somebody is watching. That is the honest answer to what a Level 2 gets you: not a purchase order, but a defensible account of how the building really runs and what that costs. See also energy efficient MEP design and operating costs.

Key Takeaways

• Audit levels are defined by ANSI/ASHRAE/ACCA Standard 211 and are cumulative. Specify the standard by name in the request for proposal.
• Benchmark first, Level 1 the outliers, Level 2 what the Level 1 confirms, and Level 3 only the measures large enough to warrant it.
• A Level 2 must include an end use breakdown accounting for 90 to 100 percent of billed energy. Check that reconciliation before reading anything else.
• Level 2 supports a capital plan and an incentive application. Level 3 supports a lender, a guarantee or a performance contract.
• Check whether measure savings are interactive or simply added, and whether the baseline is as found or as designed.
• Separate no cost and low cost measures from capital measures. The first group can be funded from operations this quarter.
• Retro-commissioning findings usually carry the shortest payback, with LBNL data putting median existing building commissioning payback at 1.7 years.
• Decide the measurement and verification approach under IPMVP before implementation, because the baseline data has to be captured first.

10. Where Budlong Works

Audit work is local in ways design work is not. Benchmarking ordinances differ by city, incentive programmes and rate schedules differ by utility, and the plan check treatment of the alteration work that follows an audit differs by jurisdiction. Budlong maintains teams across California so the incentive route and the permit route are understood before the measure list is written. More on our company story page.

Sectors where audit findings differ most

Healthcare carries ventilation minimums that remove whole categories of measure. Laboratories concentrate savings in exhaust and heat recovery, commercial offices in schedules and sequences, multifamily in central domestic hot water and common area loads.

11. Related Reading

12. Frequently Asked Questions