Building HVAC Commissioning and Energy Performance Validation Project

This project produces a commissioning and energy-performance validation package for an installed building HVAC system. The goal is not to repeat a design narrative or to solve isolated load calculations. The goal is to prove that the building delivers the required indoor service with credible energy, control, air, water, and measurement evidence.

The project is written for a real or realistic building after installation, retrofit, recommissioning, or controls upgrade. It can be applied to offices, laboratories, teaching buildings, hospitals, light industrial buildings, and mixed-use facilities. The engineering structure is the same: define the service boundary, collect evidence, test functional behavior, calculate whether measured performance is plausible, and decide whether the system can be accepted.

Project Objective

Create a reviewable HVAC commissioning package for one building or one major building zone. The final deliverable should allow an engineering reviewer to answer:

1. Which indoor service was commissioned: temperature, humidity, ventilation, pressure relationship, filtration, operating schedule, demand response, or energy target?
2. Which systems are inside the validation boundary?
3. Which instruments, BAS trends, meters, test ports, and field measurements support each claim?
4. Do air-side, water-side, electrical, and control evidence agree within expected uncertainty?
5. Which functional tests passed, failed, or need follow-up?
6. What residual risks remain before handover?

The package should be concise enough for a project closeout review and detailed enough that a future engineer can repeat the tests or diagnose drift.

What This Project Is Not

This is not a full building energy model, not an ASHRAE compliance document, and not a generic checklist. It is also not a single-fault case study. The project focuses on commissioning evidence that connects measured HVAC operation to useful building service and energy performance.

Use the companion formula sheet for equations, the exercise set for individual calculations, and the economizer case study for a focused fault diagnosis. This project combines those ideas into an acceptance package.

Scenario

Use the following scenario or replace it with site-specific data.

A three-story academic office and laboratory support building has completed an HVAC controls upgrade. The project added supply-air temperature reset, static-pressure reset, chilled-water valve tuning, economizer logic review, heat-pump plant trend points, and demand-response capability. The owner wants evidence that energy performance improved without reducing ventilation, comfort, or operational reliability.

Item	Value or requirement
Floor area	8,400\ \text{m}^2
Main AHUs	3 variable-air-volume units
Commissioned AHU for detailed test	AHU-2
AHU-2 design supply airflow	8.0\ \text{m}^3/\text{s}
AHU-2 occupied minimum outdoor-air fraction	25\%
Typical occupied return-air temperature	24^\circ\text{C}
Shoulder-season outdoor-air test temperature	12^\circ\text{C}
Supply-air temperature target during economizer test	16^\circ\text{C}
Chilled-water flow during coil validation	4.2\ \text{kg}/\text{s}
Chilled-water supply and return during validation	7.0^\circ\text{C} and 12.5^\circ\text{C}
Air temperature entering and leaving cooling coil	24.0^\circ\text{C} and 14.0^\circ\text{C}
Heat-pump thermal output during validation	145\ \text{kW}
Heat-pump electrical input inside building boundary	48\ \text{kW}
Demand-response target	40\ \text{kW} reduction for 2\ \text{h}
Comfort acceptance band	22^\circ\text{C} to 25^\circ\text{C}
Relative humidity acceptance band	35\% to 60\%
Energy-performance acceptance	measured savings plausible within uncertainty and no service violation

These numbers are deliberately compact. A real commissioning package should include seasonal tests, TAB reports, equipment schedules, sensor calibration records, sequence-of-operation documents, alarm history, maintenance constraints, local code requirements, and owner-specific acceptance criteria.

Final Deliverables

The completed package should include:

1. service boundary and baseline statement;
2. point list and trend plan;
3. instrument and calibration status;
4. pre-functional checklist;
5. air-side functional test records;
6. hydronic and coil performance reconciliation;
7. heat-pump or plant COP boundary calculation;
8. demand-response test result;
9. fault log and corrective-action table;
10. measurement uncertainty summary;
11. acceptance matrix;
12. handover memo with residual risks.

Each deliverable should be traceable. A screenshot without time stamp, sensor identity, operating mode, and boundary definition is weak evidence.

Step 1: Define the Service Boundary

Start by writing what service is being validated. For this scenario, the useful service is not “low energy” by itself. The building must provide occupied thermal comfort, ventilation, humidity control, pressure relationships where required, safe plant operation, and demand-response capability.

The validation boundary includes:

• the three main AHUs;
• outdoor-air and return-air dampers;
• cooling coils and chilled-water control valves;
• heat-pump plant electrical input inside the building meter boundary;
• VAV terminal airflow and occupied-zone temperature;
• BAS trend data and alarms;
• utility interval meter data;
• demand-response dispatch signal and response evidence.

The validation boundary excludes unrelated plug loads, lighting retrofits not controlled by the HVAC project, weather effects outside the normalization method, and process loads not served by the commissioned systems.

The baseline statement should identify the comparison period, weather basis, occupancy schedule, metering boundary, known abnormal days, and any excluded operating changes. Without this statement, an apparent energy saving may only reflect weather, occupancy, or schedule drift.

Step 2: Build an Evidence Map

Create a one-page evidence map before testing. It prevents the project from collecting impressive-looking data that does not answer the acceptance question.

Claim	Evidence required	Weak evidence to reject
Ventilation is preserved	outdoor-air fraction, VAV minimums, occupied schedule, CO2 or outdoor-air measurement when available	only a controller command
Economizer works	outdoor-air, return-air, mixed-air, supply-air, damper command, lockout state, cooling-coil valve position	only economizer enable status
Coil performance is plausible	air-side and water-side heat rates, valve position, flow, temperatures, uncertainty	only chilled-water valve percent
Heat-pump performance is credible	thermal output, electrical input boundary, operating mode, auxiliary loads	only catalog COP
Demand response is usable	dispatch time, kW reduction, zone response, rebound energy, override log	only setpoint command
Handover is safe	fault log, open issues, alarm limits, operator notes, owner acceptance	only a signed checklist

Every row should identify the data source, time stamp, owner, and pass/fail criterion.

Step 3: Design the Trend Plan

Commissioning data must be collected at a resolution that can see the behavior being tested. Five-minute utility data may be enough for a two-hour demand-response test. It is usually too slow for damper stroke behavior or unstable control loops.

For AHU-2, trend at least:

• supply airflow;
• supply fan speed and power where available;
• supply-air static pressure;
• outdoor-air, return-air, mixed-air, and supply-air temperature;
• outdoor-air damper command;
• return-air damper command if available;
• chilled-water valve command;
• chilled-water supply and return temperature;
• chilled-water flow;
• occupied mode;
• economizer enable and lockout flags;
• VAV critical-zone damper positions;
• representative zone temperatures;
• alarms and overrides.

A practical trend plan is:

Test	Minimum sample interval	Minimum duration
Damper stroke test	30\ \text{s}	one full open-close-open cycle
Economizer functional test	1\ \text{min}	45\ \text{min} stable outdoor condition
Static-pressure reset	1\ \text{min}	at least two load changes
Chilled-water coil reconciliation	1\ \text{min}	30\ \text{min} stable condition
Heat-pump COP boundary	1\ \text{min} to 5\ \text{min}	60\ \text{min} stable mode
Demand response	5\ \text{min}	event plus rebound period

Store raw data, not only plots. Reviewers should be able to recalculate the main results.

Step 4: Complete Pre-Functional Checks

Functional testing should not start until installation and point integrity are credible. A failed pre-functional check does not always stop the project, but it changes which conclusions can be accepted.

Check	Acceptance criterion	Engineering reason
Sensor identity	point name matches physical sensor and location	prevents wrong-point validation
Temperature sensors	recent calibration or field comparison within accepted tolerance	avoids false load calculation
Airflow stations	TAB or field check available for operating range	avoids false ventilation claim
Valve and damper stroke	commanded end stops match physical end stops	separates command from delivered flow
Fan and pump rotation	verified after electrical work	prevents flow and efficiency errors
BAS override log	no hidden manual overrides during test	protects interpretation
Safety and freeze protection	active and documented	commissioning must not defeat protection
Alarm routing	critical alarms visible to operators	handover depends on response path

If an item fails, record the effect. For example, an uncalibrated mixed-air sensor may still allow a damper stroke test, but it cannot support a precise outdoor-air fraction calculation.

Step 5: Test Outdoor-Air and Economizer Performance

An economizer test should prove thermal result, air path, and control state together. The following calculation uses AHU-2 during a cool shoulder-season period.

Measured stable values:

Quantity	Value
Return-air temperature, T_{RA}	24.0^\circ\text{C}
Outdoor-air temperature, T_{OA}	12.0^\circ\text{C}
Mixed-air temperature, T_{mix}	18.0^\circ\text{C}
Supply-air setpoint	16.0^\circ\text{C}
Outdoor-air damper command	63\% open
Economizer enable	active
Chilled-water valve	12\% open

For a dry sensible check:

Solve for outdoor-air fraction:

Substitute:

The inferred outdoor-air fraction is 50\%. The outdoor-air fraction required to reach the supply-air target before mechanical cooling is:

Engineering Interpretation

The economizer is enabled and outdoor air is above the minimum ventilation fraction, but it is not providing enough outdoor air to meet the supply-air setpoint without mechanical cooling. That does not automatically prove a fault. The commissioning team must check humidity lockout, freeze protection, minimum discharge-air limits, damper authority, actuator stroke, building pressure constraints, and relief-air capacity.

If none of those constraints applies, the acceptance result is conditional fail: the economizer sequence or air path does not deliver the available cooling service. The test should be repeated after correcting damper linkage, actuator span, relief path, or sequence limits.

Step 6: Reconcile Air-Side and Water-Side Coil Heat

Cooling-coil validation is stronger when the air-side and water-side heat rates agree. Use stable data after sensors are checked.

Water-side heat rate:

With \dot m_w=4.2\ \text{kg}/\text{s}, c_{p,w}=4.18\ \text{kJ}/(\text{kg}\cdot\text{K}), T_{w,return}=12.5^\circ\text{C}, and T_{w,supply}=7.0^\circ\text{C}:

Air-side sensible heat rate:

With \rho_a=1.2\ \text{kg}/\text{m}^3, \dot V_a=8.0\ \text{m}^3/\text{s}, c_{p,a}=1.006\ \text{kJ}/(\text{kg}\cdot\text{K}), T_{a,in}=24.0^\circ\text{C}, and T_{a,out}=14.0^\circ\text{C}:

Engineering Interpretation

The two independent heat-rate estimates agree because the example was selected as a clean validation case. In a field test, agreement within a stated uncertainty band is the objective. A large mismatch can indicate airflow error, water-flow error, sensor offset, unmeasured latent load, coil bypass, unstable conditions, stratified mixed air, or incorrect point mapping.

The commissioning package should not hide this reconciliation. If only the water side is measured, state that latent and air-distribution performance were not independently verified.

Step 7: Check Fan Reset and Airflow Energy

Variable-air-volume systems should reduce fan power strongly when airflow falls, but only if dampers, static-pressure reset, and duct pressure control allow it.

Assume AHU-2 fan power is 18.0\ \text{kW} at 8.0\ \text{m}^3/\text{s}. During partial load, airflow falls to 6.8\ \text{m}^3/\text{s}. A first screening with the fan affinity relation gives:

The trend shows 13.8\ \text{kW} instead.

Engineering Interpretation

The measured value is not automatically wrong. Real systems have static-pressure limits, minimum airflows, filter loading, duct leakage, VAV damper positions, and fan efficiency changes. However, the gap is large enough to review critical-zone damper positions and the static-pressure reset sequence. If most VAV dampers are partly closed while static pressure remains high, the fan is doing avoidable throttling work.

Acceptance should not require exact cube-law behavior. It should require that the reset sequence reduces pressure when zones can accept it and that the owner receives a trend-based baseline for future drift.

Step 8: Validate Heat-Pump COP Boundary

Heat-pump performance depends strongly on the chosen electrical boundary. A manufacturer compressor COP is not the same as building meter COP.

Measured during a stable heating test:

Quantity	Value
Useful thermal output to building loop	145\ \text{kW}
Compressor input	41\ \text{kW}
Heat-pump fans, controls, and local pumps	7\ \text{kW}
Electrical input inside validation boundary	48\ \text{kW}

Compressor-only COP:

Building-boundary COP:

Engineering Interpretation

Both numbers can be correct, but only the second matches the boundary used for building energy validation. The acceptance memo should state which COP is used and why. If utility savings are claimed, include auxiliary power that appears inside the metered boundary.

This distinction also protects future troubleshooting. A plant may have acceptable compressor performance but poor system performance because pumps run continuously, valves leak, defrost strategy is excessive, or controls force simultaneous heating and cooling.

Step 9: Run a Demand-Response Validation Test

Demand response should reduce power without violating service constraints. For this scenario, the building receives a two-hour event request.

Observed event result:

Quantity	Value
Baseline HVAC electrical demand	286\ \text{kW}
Average demand during event	244\ \text{kW}
Event duration	2.0\ \text{h}
Maximum occupied zone temperature	24.7^\circ\text{C}
Relative humidity range	42\% to 55\%
Rebound energy above baseline after event	18\ \text{kWh}

Average demand reduction:

Event energy reduction before rebound:

Net event energy reduction after rebound:

Engineering Interpretation

The 40\ \text{kW} reduction target was met, and the maximum zone temperature remained inside the 22^\circ\text{C} to 25^\circ\text{C} comfort band. The rebound is not a failure, but it must be reported because it affects energy, capacity planning, and repeated-event readiness.

The test is acceptable only if ventilation and pressure constraints were also preserved. A demand-response strategy that saves power by disabling required outdoor air or hiding alarms is not acceptable commissioning evidence.

Step 10: Estimate Measurement Uncertainty

Acceptance criteria should be wider than measurement noise but tighter than the engineering risk allows. A simple uncertainty screen is often enough for a commissioning package.

For the air-side coil heat rate:

Assume:

• airflow uncertainty: 5.0\%;
• air temperature difference uncertainty: 0.4\ \text{K} over a 10.0\ \text{K} difference, or 4.0\%;
• density and heat-capacity screening uncertainty: 2.0\%.

Approximate combined relative uncertainty:

The relative standard uncertainty is about 6.7\%. A practical commissioning acceptance band might use a wider action band, such as \pm10\%, to allow for field stability and sensor placement.

Engineering Interpretation

Uncertainty does not excuse poor data. It helps the reviewer distinguish a meaningful performance failure from normal measurement scatter. If two heat-rate estimates differ by 4\%, the result is probably acceptable. If they differ by 25\%, the package should identify the dominant error source before handover.

Step 11: Create the Acceptance Matrix

Summarize results in a decision matrix. Avoid vague statuses such as “looks good.”

Requirement	Evidence	Result	Disposition
Occupied comfort maintained	zone trends during normal and demand-response tests	pass	accept with seasonal retest
Minimum ventilation preserved	outdoor-air fraction and VAV minimum checks	pass with two sensor caveats	calibrate sensors before final seasonal trend
Economizer provides free cooling when available	mixed-air calculation and damper stroke test	conditional fail	correct actuator span and repeat
Coil heat rates reconcile	air-side and water-side calculations	pass	store raw data and calculation sheet
Fan reset reduces avoidable pressure	fan trend and VAV damper review	partial pass	tune critical-zone logic
Heat-pump COP boundary is documented	thermal output and electrical input boundary	pass	report boundary COP, not compressor-only COP
Demand response meets target	interval meter, BAS trends, rebound review	pass	approve for limited events
Handover evidence is usable	point list, trend files, fault log, owner memo	partial pass	close open issues before warranty turnover

A conditional fail is not a paperwork failure. It is useful engineering information. The team should record corrective action, owner risk acceptance if any, and the required retest.

Step 12: Write the Handover Memo

The final memo should be short and explicit:

1. state the commissioned boundary;
2. list the tests performed and dates;
3. summarize pass, partial pass, fail, and deferred items;
4. attach raw trend data or identify storage location;
5. identify calibration assumptions;
6. list unresolved issues and owners;
7. state operating limits until retest;
8. define seasonal retest requirements;
9. name the accepted baseline for future fault detection;
10. record owner signoff or engineering hold point.

Do not bury major failures in an appendix. If the economizer failed, the first page should say so.

Common Failure Modes

Building HVAC commissioning often fails because the package validates commands instead of delivered service. Common failure modes include:

• BAS point names do not match physical sensors;
• mixed-air sensors are stratified or poorly located;
• outdoor-air dampers move but blades do not reach commanded position;
• VAV minimums are overridden during energy tests;
• static-pressure reset is defeated by one bad critical-zone signal;
• chilled-water valve position is treated as cooling load;
• heat-pump COP excludes auxiliary loads inside the project boundary;
• demand-response tests ignore rebound and comfort recovery;
• fault logs are cleared before engineering review;
• seasonal tests are promised but never assigned to an owner.

Each failure mode should map to a preventive control: calibration, physical stroke test, trend review, alarm ownership, retest criterion, or explicit residual-risk acceptance.

Review Checklist

Before closing the project, confirm:

1. the service boundary is written and matches the acceptance claims;
2. all trend points have names, units, sampling intervals, and time ranges;
3. each calculation has units and a plausibility check;
4. air-side and water-side evidence are reconciled where possible;
5. ventilation and comfort were not sacrificed for energy savings;
6. demand response includes rebound and override review;
7. uncertainty is stated for measured performance claims;
8. failures have corrective actions and retest owners;
9. the final memo distinguishes accepted, conditional, deferred, and rejected claims;
10. the owner can use the package as a baseline for future fault detection.

The value of the project is not the number of checklist items completed. The value is that a future engineer can see what the building was supposed to do, how the evidence was collected, which calculations supported acceptance, and which risks remained after handover.