How to Specify and Verify an HTM-03-01 Hospital AHU

Updated August 2026

An HTM-03-01 Hospital AHU is an air-handling unit specified as part of a healthcare ventilation system governed by HTM 03-01. It cannot be accepted from a product label or filter grade alone. For healthcare projects in England, the ventilation brief must connect clinical use and room risk to airflow, pressure, distribution, controls, commissioning, independent validation, handover, and ongoing verification.

This is an England-focused decision guide, not a substitute for the current NHS documents, project risk assessment, or an Authorising Engineer (Ventilation). Scotland publishes SHTM 03-01 and Wales publishes WHTM 03-01. Core English guidance remains the 2021 Part A and Part B pair, while NETB 2023/01A is an addendum for portable and semi-fixed HEPA air-cleaning devices.

What Does HTM 03-01 Cover, and How Do Part A and Part B Work Together?

HTM 03-01 is a Health Technical Memorandum for specialised ventilation in healthcare premises. For new installations and major refurbishments, Part A sets the design and validation standard. During operation, Part B governs maintenance, management, and performance verification. Neither health technical document turns a standalone AHU into a compliant installed system.

On its HTM 03-01 Part A document, NHS England identifies the 2021 edition as the design and validation reference for new installations and major refurbishments. The companion Part B governs operational management and performance verification. A live publication check still matters at the start of every project because an addendum can extend the evidence set without replacing the core edition.

Is HTM 03-01 a legal requirement?

HTM 03-01 is NHS technical guidance, not a single Act of Parliament. Its language distinguishes requirements expressed as “must” from recommendations expressed as “should.” That doesn’t make it optional in practice: NHS projects, organisational policies, contracts, risk controls, and statutory duties can make compliance evidence central to acceptance. HSE separately notes the legal duty to provide sufficient fresh or purified air in enclosed workplaces. A safer question is: “What risk, contractual, policy, and legal basis supports any alternative?”

Paragraphs 4.10 and 4.11 require a proposed derogation or alternative strategy to receive scrutiny, written VSG agreement, and evidence that safety is no less. Evidence that a method has merely been used elsewhere isn’t enough. Record the reason, limits, approver, supporting evidence, and revalidation trigger.

Start With the Clinical Function, Not the AHU Catalogue

The first specification is a clinical and operational brief, not a model number. The scope-and-governance framework above becomes usable only when the clinical team turns it into room-specific inputs. Define who uses the room, what the procedure releases, which direction contamination must move, how long the space operates, what happens on failure, and who owns the system after handover.

A room name alone can be ambiguous. “Isolation,” “treatment,” or “critical care” doesn’t automatically identify the applicable schedule row or pressure strategy. HSE’s general ventilation risk guidance also notes that adequacy depends on factors such as occupancy, floor space, and work activity. HTM 03-01 Part A still requires VSG and clinical stakeholders to agree the room classification, patient group, procedure, adjacent-space relationship, operating mode, and failure response before the mechanical schedule is frozen.

Once those facts are agreed, product selection can begin. Koven Air’s HTM 03-01 compliant hospital AHU solution page covers equipment and the project-configuration route. This guide deliberately stops short of model comparison, prices, and quotations so the two pages serve different search intent.

Translate Room Risk Into Airflow, Pressure and Air-Distribution Logic

Air changes per hour are one output, not the design method. Those clinical inputs become the boundary conditions for the airflow calculation. Under Part A paragraph 4.63, minimum supply is the greatest of five demands: fresh air, pressure or open-door need, thermal load, desired air-change rate, and make-up air for local extract. Each served room needs a scoped calculation.

Pressure is also relational. The negative room must be negative to a named adjacent space under defined door and plant conditions; a positive room must protect a named cleaner zone. Supply and extract quantities, fabric leakage, door movement, transfer paths, controls, and alarms work together. Fan duty alone doesn’t prove direction.

The table is a navigation aid, not a substitute schedule. Values can depend on sub-room, lobby, supply-air category, distribution pattern, local extract, operational mode, and other notes in the source. Koven Air’s clinical room ventilation lookup helps identify candidate rows; verify each one against the official document and clinical brief.

Worked airflow example: calculate, then qualify

Suppose a 45 m² isolation room has a 3 m clear height. Its room volume is 135 m³. A schedule value of 10 AC/h corresponds to 1,350 m³/h. That is only the air-change candidate. Designers must still compare it with fresh-air demand, pressure/door-opening duty, cooling or heating load, and local-extract make-up. The largest governs under paragraph 4.63.

In this context, paragraph 4.68 calls for 5% for leakage and balancing plus 5% for uncertainty, a combined 10% design margin. Maximum duct leakage in the cited test is 3%. Applied illustratively to 1,350 m³/h, a 10% margin gives 1,485 m³/h, but this must not be mistaken for the final fan selection: system resistance, filter condition, control range, redundancy, heat recovery, sound, and actual scheduled notes remain unresolved.

For early arithmetic, Koven Air’s hospital AHU airflow sizing estimator can organize inputs. Treat its output as a preliminary check and preserve the approved design calculation as the acceptance record.

Air distribution determines whether the calculated air reaches the risk

Room-average AC/h can’t show what happens between a patient, clinician, contaminant source, supply terminal, door, and extract grille. Record the intended clean-to-less-clean path and inspect it under the real room layout. Tall equipment, privacy curtains, lights, pendants, storage, furniture, and open doors can redirect or short-circuit a theoretically adequate airflow.

Use a method suited to the question. Airflow measurements establish quantity; differential-pressure measurements test the relationship between named spaces; smoke visualization can show direction and local behavior; application-specific velocity tests address systems such as ultra-clean ventilation. Capture terminal positions, door states, control mode, occupancy simulation, instrument identity, calibration, and environmental conditions so the result can be repeated.

Pressure stability must be checked across the operating range, not only at one balancing point. Test clean and loaded filter states where the design requires them, normal and standby fan operation, relevant door movement, and the response to an alarm or loss of extract. A system that meets −5 Pa with every door closed but loses the intended direction during normal clinical movement has exposed a control or fabric problem, not merely a record-keeping issue.

Filtration, Drainage and Cleanability Must Be Verified as an Installed System

Filter class describes media performance under a test framework; it does not prove the housing is sealed, the filter is seated, the access route is safe, the drain works, or room air reaches the filter effectively. Specify component, installation, test, maintenance, and disposal evidence together.

For filtration, Part A paragraph 4.30 gives an important correction to equipment-first thinking: almost all viable particles in a ventilated room originate from occupants rather than incoming air, so dilution is generally more important. ISO ePM1 ≥50% is cited for general areas; EPA/HEPA supply filtration is normally reserved for ultra-clean and designated clean rooms.

Special extract filtration is also risk based. Paragraph 9.61 says EPA/HEPA may be used where hazardous substances or organisms could be discharged, but the assessment must address need, installed validation, safe filter changing, and disposal. A general mortuary or post-mortem extract does not automatically require filtration.

NETB 2023/01A does not overturn the risk-based message. It addresses local portable and semi-fixed HEPA devices, recommends H13 or H14 performance for healthcare applications, and warns that air cleaners must not be used as a reason to reduce fresh-air ventilation. Room performance depends on airflow rate, device placement, room layout, background ventilation, operation, and maintenance, not filter efficiency alone.

In a healthcare environment, a stringent air-quality target still needs a defined engineering purpose. Healthcare facilities and healthcare buildings contain different clinical risks, room geometries, occupants, and operating patterns; NHS Improvement-era labels or older project documents should never replace the current source set and locally approved risk assessment.

Cleanroom projects can introduce a different classification and validation context. The cleanroom HVAC validation guide explains that adjacent topic; it should not be used to replace the healthcare room schedule.

Integrate the AHU With Plant Space, Controls, Resilience and Maintenance Access

Even a suitable AHU can fail its clinical purpose when the plantroom, duct system, controls, drainage, power, or maintenance route is wrong. The same installed-system logic that governs filtration and drainage also applies to the surrounding plant and controls. Integration review must prove safe installation, testing, isolation, cleaning, repair, and replacement under the agreed operating and failure strategy.

Start at the air intake and finish at the discharge, as the validator will. HSE’s mechanical-ventilation guidance points to the full duct-and-fan route and the need for an engineer to check whether a system provides adequate ventilation. The healthcare-specific whole-system validation route in HTM 03-01 Part A then means checking contamination sources around the intake, weather protection, access, fan and filter removal, coil withdrawal, drain gradients, trap pressure, duct cleanability, measurement stations, fire interfaces, terminal distribution, exhaust discharge, and the route for safe filter disposal.

Part A paragraph 9.12 calls for airflow nature and direction to be clearly marked on ducts. Paragraph 9.19 specifies motorised, spring-return, low-leakage BS EN 1751 class 3 isolation dampers at AHU connections, closing on failure or shutdown to prevent reversal. These details turn a drawing arrow into a testable failure state.

Controls need an evidence owner

For every alarm and mode, name the initiating condition, sensor, setpoint or threshold, delay, commanded response, displayed message, recipient, escalation path, and reset authority. Test normal, setback, fire, purge, isolation, loss of airflow, dirty filter, fan failure, sensor failure, power restoration, and duty/standby changeover where applicable.

Measurement facilities should be designed in, not improvised later. Accessible duct test points, calibrated instruments, reliable differential-pressure taps, room reference points, and BMS trends make annual verification repeatable. CIBSE’s healthcare measurement discussion reinforces the practical point: inaccessible or poorly located sensors can make a compliant intent unprovable.

Resilience is a clinical consequence decision

Don’t apply redundancy as a generic badge. Define the consequence and permitted duration of losing supply, extract, pressure cascade, temperature, humidity, or monitoring. Then decide whether the response needs standby fans, dual power, automatic changeover, local indication, remote alarm, protected controls, or an operational shutdown plan. Record what the system can maintain during maintenance, not only during an electrical fault.

Existing plant also needs an honest suitability check. Retention under Part A is conditional and uses a 10-year age threshold alongside the ability to achieve full compliance. Midlife refurbishment under Part B often occurs at about 10 years, with the system taken out of use, inspected, cleaned, assessed for corrosion, considered for controls upgrades, and validated before return. Age alone neither condemns nor validates the plant.

Commissioning Proves Performance; Independent Validation Proves Fitness for Purpose

Commissioning brings the installed system to its specified operating condition and records performance. With integration resolved, commissioning can test the installed operating state. Independent validation asks whether the whole system is fit for its agreed healthcare purpose and whether the client should accept it. The second cannot be replaced by the contractor’s commissioning sheet or an AHU factory test.

Commissioning guidance in paragraph 11.3 describes the process as essential and expects provision at the design stage, referring to relevant CIBSE codes and BSRIA BG49. Paragraph 11.27 says construction should be complete before dynamic commissioning: rubbish removed, floors cleaned, panels and access complete, and ceilings finished. Otherwise dust, missing fabric, open panels, and unfinished doors can invalidate the measured state.

Paragraphs 12.1 to 12.4 then establish independent validation before client acceptance for new and refurbished systems. The client appoints a suitably qualified engineer—an AE(V) or similarly qualified person—who is independent of the designers, contractors, suppliers, installers, commissioners, and later operators. Validation covers the entire system from intake to discharge.

“All new and refurbished ventilation systems should be independently validated prior to acceptance by the client.”

Build a witnessable test sequence

1. Confirm readiness: construction, cleaning, access, labels, doors, ceilings, controls, utilities, and calibrated instruments are complete.
2. Inspect the system: trace intake, AHU sections, ducts, dampers, terminals, rooms, extracts, and discharge against drawings.
3. Prove functions. Test modes, interlocks, failure actions, alarms, standby operation, BMS points, and safe reset.
4. Measure performance. Record main and branch airflows, room supply/extract, air-change rate where applicable, pressure relationships, temperature, humidity, noise, and application-specific velocity.
5. Resolve deviations. Link each defect to an owner, corrective action, retest, and acceptance status.
6. Issue the conclusion. Paragraph 12.32 requires a full report and a clear statement on whether the agreed design standard was achieved.

Handover Creates the Evidence Baseline for Part B Governance

Handover is the transfer of a measurable operating baseline, not a box of manuals. The validation conclusion then becomes the starting point for operation, rather than a one-off acceptance record. Operations teams need the design, installation, measurements, accepted settings, outstanding conditions, and recurring-task owners in one controlled record.

Part A paragraph 11.11 requires commissioning information to be assembled in the commissioning manual and handed over with as-fitted drawings. Paragraph 13.3 expands the information baseline, including drawings, schematics, and room design-versus-actual airflow data. Paragraph 13.36 requires training for end users and for those who operate and maintain the system.

Governance under Part B assigns roles rather than leaving “facilities” as an unnamed owner. A healthcare organisation establishes a Ventilation Safety Group linked to board-level governance. AE(V) provides independent audit and advice; AP(V) implements the safety policy in practice; and CP(V) performs maintenance and periodic testing within competence and appointment limits.

Record retention is measurable. Paragraph 1.44 calls for at least five years for ordinary ventilation records and 25 years for manufacturing-pharmacy records. Paragraph 5.46 also sets at least five years for maintenance records. Retention needs a named repository, document owner, version method, access control, backup, and link to the system inventory.

Inspection, Verification, Failure Response and Revalidation Under Part B

Inspection under Part B separates frequent observation from detailed measurement. All ventilation systems receive at least a simple annual visual inspection. Critical healthcare ventilation is visually inspected quarterly and performance-verified annually. AHUs receive internal and external visual inspection at least every three months. These intervals don’t remove the need for additional risk-based checks.

Part B paragraph 1.47 makes the distinction explicit: visual inspection checks apparent condition; annual verification measures and records performance. Paragraphs 5.6 and 5.7 add AHU condition, connections, controls, the ability to maintain conditions, and drainage cleanliness and operation. Tray cleanliness and water colour can expose a fault.

What happens when critical ventilation fails verification?

Paragraph 4.34 says a failed critical system shouldn’t be returned to service as though nothing happened. It requires the duty manager to be informed, with reporting to the user department, infection prevention and control, and the AP(V). Next, the organisation controls the clinical risk, investigates the cause, records decisions, repairs or modifies the system, and repeats the required evidence.

After refurbishment, paragraph 4.35 requires full validation before reuse. A passed component test or restored fan status isn’t necessarily enough: the acceptance question concerns the served clinical environment and whole-system function. Define who can declare the system unavailable, who can accept temporary operational controls, and who can authorize return to clinical use.

The Clinical Risk-to-Evidence Chain: A Five-Gate Acceptance Check

Koven Air’s Clinical Risk-to-Evidence Chain is a decision framework derived from the lifecycle in HTM 03-01 Part A and Part B; it isn’t a replacement standard. A project advances only when each gate has an accountable owner, an approved input, measurable evidence, and a clear disposition for open items. Failure at one gate remains visible.

This framework prevents three common handoff errors. First, it stops a product certificate from being stretched into system acceptance. Second, it stops successful commissioning from being labelled independent validation. Third, it stops a passed validation from being treated as permanent: ongoing governance must maintain and periodically re-prove the accepted condition.

Use the gates during bid comparison

A bid comparison should score evidence against the same approved brief. Ask every bidder to state assumptions, exclusions, calculation conditions, filter stages, casing and leakage basis, fan operating points, spare capacity, sound basis, control interface, access envelope, drainage arrangement, commissioning scope, validation support, handover records, and training. Separate “included,” “by others,” and “not demonstrated.”

Don’t award an apparent technical advantage that exists only because one proposal assumed a lower airflow, cleaner filter, smaller external static pressure, easier sound criterion, or narrower controls scope. Normalize the inputs first. When an alternative strategy is proposed, require the safety evidence, limits, VSG route, and future verification method before treating it as equal.

This result is a procurement handoff without turning the guide into a supplier-comparison page. The clinical brief and acceptance gates define what evidence a product solution must provide; the commercial page can then present configuration, drawings, selection data, and quotation against those frozen inputs.

Frequently Asked Questions