Industrial HVAC systems work hard, often around the clock and under conditions that commercial offices never see. Dust-laden intake air, high internal heat loads, variable production schedules, and the sheer scale of ductwork and plant complicate both operation and compliance. A TM44 inspection, carried out under the UK framework for air conditioning energy assessments, can be dismissed as a checkbox exercise. Done properly, it becomes a practical roadmap for lowering energy use, improving resilience, and avoiding unplanned downtime. The differences come from preparation, field technique, and how findings are translated into action.
What follows draws on site inspections across factories, data halls, and process environments where comfort cooling and critical cooling coexist. TM44 provides the structure: evaluate system documentation, inspect equipment and controls, review maintenance, and recommend cost-effective improvements. The best results come from treating that structure as a floor, not a ceiling.
Framing TM44 in an industrial context
TM44 inspections apply to air conditioning systems with an effective rated output greater than 12 kW, including multiple units that serve the same building. Many industrial sites cross this threshold easily, even if the plant is distributed. Unlike offices, where systems are comparatively homogeneous, industrial sites often hide cooling capacity in unusual places. Packaged units on mezzanines, DX splits serving plant rooms, ad hoc temporary units feeding critical process areas, and legacy chillers running because nobody wants to risk turning them off. The TM44 assessor’s first job is to build a map of the actual cooling system as it operates today, not as it was designed a decade ago.
The pressing reasons go beyond compliance. Energy intensity on industrial cooling is high, often representing 20 to 40 percent of total electricity use on heat-heavy processes. Poorly tuned controls or obstructed heat exchangers can add five to fifteen percent to consumption with no visible symptom apart from higher bills. Good TM44 practice finds these inefficiencies, quantifies them, and relates them to plant risk.
Preparation that pays off
Sites can set themselves up for a meaningful TM44 outcome by getting documentation and access in order before the visit. The assessor’s time on site is limited, and every minute spent hunting for an air handling unit behind a locked cage is a minute not spent diagnosing real savings.
A short pre-visit checklist keeps things smooth:
- Accurate site plan of cooling plant: chillers, packaged units, splits, AHUs, CRAC/CRAHe units, cooling towers, pumps, and major duct routes. Recent BMS screenshots or trend extracts showing typical weekly operation, with setpoints, schedules, and alarms. Maintenance logs for the last 24 months: F-gas records, leak tests, filter changes, coil cleaning, oil analysis where applicable, and any major repairs. Refrigerant inventory by circuit, with type and charge, plus any retrofit or top-up records. Health and safety constraints: induction requirements, working at height needs, confined spaces, hot work permits, and isolation procedures.
Two clarifications are especially helpful. First, identify controlled environments with non-negotiable temperature or humidity requirements, and state the ranges, not just the nominal setpoint. Second, flag any areas where ad hoc temporary units have been installed. Temporary becomes permanent more often than people like to admit.
Building an operational map
On site, the assessor should begin with a walk-through that aligns the drawing set, asset register, and what is actually installed. In practice, these never match perfectly. Expect to find renamed AHUs, decommissioned splits still powered, or a supplementary chiller tied into a process loop. Photograph labeling and serial plates. Cross-check nameplate duties against the asset list. Note economiser dampers, heat recovery wheels, and VSD presence rather than relying on design intent.
Industrial airflow routes are often long and tortuous. Duct insulation can be missing at plant penetrations, and flexible sections near machinery can collapse and spring back depending on pressure, which skews measured airflow. Where supply and extract are not balanced, negative pressure can pull in warm, contaminated air through doors or service penetrations. This usually shows up as higher coil loading and dirtier filters, plus complaints about drafts on certain shifts.
An operational map should connect equipment to zones, schedules, and control logic. For example, a packaged unit on a roof might be controlled by a local thermostat, but the AHU feeding the same zone might be on the BMS, with a static schedule. If both cool simultaneously without coordination, you pay twice for the same degree of cooling.
Documentation that actually matters
TM44 requires a review of available documentation, but the volume is less important than the signal within it. A thick O&M manual rarely explains why the chiller short cycles every afternoon. Several documents, however, directly influence energy and reliability:
- Sequence of operation, ideally as-implemented, not as-designed. This explains staging, economiser logic, humidification and dehumidification priorities, and alarm responses. If there is no written sequence, the BMS point list and trend data should be used to reconstruct one. F-gas and leak logs with dates and quantities. Repeated small top-ups, especially on the same circuit, point to microleaks that quietly erode efficiency. Refrigerant retrofit records. A legacy R22 system long since retrofit to R407C or R438A may run but at compromised capacity or efficiency due to mismatched components and controls. That context shapes recommendations. Maintenance scope and intervals. Coil cleaning “as needed” often means not often enough in dusty plants. Evidence of proactive cleaning correlates with lower condensing temperatures and compressor run times. Commissioning and recommissioning reports following major changes. Many sites change fans or VSDs, then never tune setpoints or control loops to match.
If the documentation is thin, that in itself is a finding. A practical TM44 report will recommend targeted documentation that reduces risk: a concise sequence of operation, a current refrigerant inventory, and a single-page plant map.
Field techniques that separate good from average
The heart of a valuable TM44 inspection is direct observation. The assessor will not perform intrusive testing, but there is plenty to learn from accessible measurements and sensory checks. Aim to gather enough evidence to confirm how the system behaves over time, not just at the instant you are present.
Start with condenser and evaporator coil condition. Industrial air is often full of fine particulates that pass through coarse filters and paste themselves onto fins. A coil that looks “gray” to the eye can still impose a temperature penalty of several Kelvin. Infrared thermography, used judiciously, helps reveal uneven heat rejection or blocked sections. Where access is poor, a simple test using contact probes on liquid and suction lines helps infer performance.
Fan belts deserve more attention than they get. Slight slippage rarely squeals but quietly drops static pressure and airflow, which pushes compressors to lower suction pressures and longer run times. In VSD-equipped fans, a worn belt shows up as higher drive frequency to hit the same setpoint. That extra few Hertz eats power.
Refrigeration circuits should be checked for oil staining and corrosion around flare fittings and microchannel coils. UV dye use is not part of TM44, but a visible inspection picks up risk. A recurring pattern in industrial sites is elevated head pressure due to either a partial blockage of fin surface or a control setpoint left at a conservative default. Raising condensing temperature by just 5 K can push compressor power up by around 8 to 10 percent on many systems. Where safe and permitted, confirm condenser fan staging or VSD head pressure control settings during the visit.
Economiser operation is another fertile area. Plate heat exchangers or rotary wheels that should enable free cooling often sit disabled after a winter alarm that nobody investigated. Dampers can be mechanically stuck even while the BMS shows them at 50 percent open. A quick visual and a manual damper exercise can expose the truth. Logging outside air temperature against economiser status over a week, if available, adds weight to a recommendation.
Controls are usually the largest lever. Seek out evidence of setpoint drift. Spaces marked to be held at 23 degrees might run closer to 20 or 26 in reality, not because of design constraints but due to poorly located thermostats, failed sensors, or a forgotten manual override. Look for deadbands too narrow to prevent short cycling, simultaneous heating and cooling at AHUs, or multiple subsystems arbitraging each other. In a process environment, you might find that humidity targets are tighter than necessary for product quality, or that the target remains fixed year-round. Loosening a humidity band from 45 percent ±3 to ±5, if acceptable, often saves both reheat and cooling energy.
Using trends to move beyond snapshots
A common limitation of TM44 is that the assessment occurs on a single day. Industrial loads are cyclical, seasonally and by shift. The assessor should request trend data at least one week long, and preferably across shoulder and peak seasons. If the BMS lacks trends, a cost-effective fallback is to place three or four portable temperature and humidity loggers in representative locations for a week before the visit.
From trend data, several patterns emerge. First, night setback. Many sites leave cooling at day setpoints throughout the night, either because the BMS schedule is static or because a critical room in the same zone demands constant conditions. Breaking out the critical room onto its own schedule or adding local transfer air capability can allow the rest of the zone to relax, without compromising product safety. Second, staging. Watch for compressors or chillers that do not unload smoothly, which indicates a sequencing issue. Third, free cooling. Plot outside air temperature against chilled water valve positions and economiser status. If the economiser rarely engages, ask why.
Part-load operation is the default in most modern plants due to VSDs and variable loads. Efficiency curves are not linear. Chillers with magnetic bearings shine at low load, while some older screw machines become very inefficient when run lightly for extended periods. Trends reveal whether a piece of equipment sits in its sweet spot. If it never does, the TM44 report should describe a control change or, for capital planning, suggest rightsizing at end of life.
Airside compromises in industrial spaces
Production spaces suffer from dust, oil mists, and variable heat sources. Filters clog sooner than predicted, coils foul quickly, and systems spend most of their effort dealing with non-thermal burdens. A few practices consistently help.
Filter selection should be tuned to the contaminant profile. Upgrading from a coarse prefilter to a higher MERV or ISO ePM class without adding face area can choke airflow and raise fan energy. The better approach is staged filtration with higher surface area, plus a cleaning regime that prevents oil carryover into the coil. Maintenance logs tell whether the filter change interval is based on differential pressure, not guesswork.
Supply air distribution matters more than people think. High-bay spaces often rely on throw rather than ducted grilles. As processes move, airflow patterns no longer match heat sources. Large destratification fans, properly controlled, can cut winter heating demand and help cooling distribution. A TM44 inspection should comment on air throw, return locations, and signs of short-circuiting between supply and extract.
Where processes demand localized capture, such as welding or cutting, make-up air needs to be balanced to avoid negative pressure that drags in untreated air through doors. A quick smoke test near doors and penetrations reveals unwanted infiltration that increases latent load and contaminant ingress.
Water-cooled plant, towers, and hygiene
Many industrial sites rely on evaporative cooling, either through cooling towers or adiabatic systems. TM44 does not replace water hygiene regimes, but it should consider how water-side practices impact energy and resilience.
Cooling towers, in particular, reward careful observation. Drift eliminators, fan pitch and cleanliness, basin condition, and fill integrity affect approach temperature and fan power. Scale buildup steals heat exchange area and elevates condensing temperatures. Water treatment records should show stable cycles of concentration tailored to the local water chemistry. If blowdown appears excessive, the report should calculate the water and energy penalty of running at low cycles and recommend a review with the water treatment provider.
Tower fan control should be coordinated with condenser water setpoints. Many towers run conservative high setpoints year-round. Allowing lower condenser water temperatures in cooler months, within chiller manufacturer limits, increases chiller efficiency. Variable tower fan speed helps hold the target without on-off cycling that wastes power and shortens motor life.
Adiabatic systems often carry restrictions due to water use and hygiene risk. A well-run adiabatic cooler that only wets media when dry-bulb alone cannot meet setpoint provides near-free seasonal savings. If logs show almost constant wetting, controls need attention.
Refrigerants, leakage, and future-proofing
F-gas regulation continues to tighten, and the economic impact of leaks is stark. A large package or chiller with an HFC charge can cost several thousand pounds to top up after a moderate leak, and the energy penalty of running undercharged is measurable. The inspection should overlay refrigerant type, age, and leak history onto risk and energy.
Systems running on high GWP refrigerants with recurring leaks warrant a strategic recommendation: either invest in reliable leak detection and proper repair, or plan for retrofit or replacement with lower-GWP alternatives at the next major failure. This is not just an environmental stance. Many low-GWP blends also offer improved thermodynamic performance when paired with suitable compressors and controls. The decision hinges on compatibility, oil type, and safety classification. A TM44 report should not prescribe a specific refrigerant, but it can lay out decision criteria and a timeframe.
Microchannel condensers, common on modern packaged units, respond poorly to patch repairs. Once corrosion starts, leaks often recur. Where corrosion is advanced, the best practice is to protect with proper coatings early in life or accept a shortened lifespan and plan capital accordingly. Again, TM44 can TM44 document condition and trajectory, which helps plant managers avoid surprise failures in peak season.
Controls, sequencing, and human factors
Industrial sites accumulate layers of control. A legacy local thermostat sits alongside a BMS retrofit, plus a process interlock installed by a machine vendor. None of these talks to the others. TM44 calls for a review of controls that finds duplication and conflict. It also encourages actions that befit the building’s operational reality rather than theoretical best practices.
Simple changes often yield outsized results:
- Standardize and document setpoints and deadbands for cooling, heating, and humidity by zone, with justification tied to process needs. Avoid setpoint fights between systems. Enable night and weekend setbacks wherever the process allows, with exception schedules for runs and cleaning shifts. Trend and verify. Sequence chillers and compressors to stay within efficient load ranges. If two machines each at 30 percent are far worse than one at 60 percent, adjust staging thresholds. Align economiser or free cooling logic with real sensor values and functional damper feedback. If sensor drift is suspected, recalibrate or replace. Remove manual overrides that have become permanent and hide problems.
Equally important are people and permissions. If a line supervisor can lower a thermostat to 18 on hot days, they will. This is less a scolding than a design problem. Provide a local comfort increase control with limited range or a timed boost, and record those actions so the system can learn. The assessor should recommend user-level guardrails and feedback mechanisms that keep the system within planned operation without frustrating occupants or operators.
Maintenance that protects efficiency
A TM44 inspection must review maintenance practices and their impact on energy. Many activities are performed for reliability or hygiene, yet they also influence consumption. It helps to present maintenance recommendations framed in cost, frequency, and benefit.
Coil cleaning, when done with the right chemistry and technique, pays back fast. The energy penalty of a dirty coil scales with airflow and ambient temperature. After cleaning dozens of industrial condensers, I have seen head pressures drop by 10 to 20 percent and compressor amperage follow. Schedule coil inspections and cleanings based on pressure and temperature deltas, not fixed dates alone. For AHUs, monitor coil pressure drop and temperature approach alongside filter differential pressure.
Belt inspection and tensioning are undervalued. A belt drive that slips just slightly erodes fan efficiency by several percent. Converting to direct drive with VSD at end of life makes sense in many cases, but until then, proper alignment and tension matter.
Refrigerant circuit health should include periodic subcooling and superheat checks where accessible and safe. If readings deviate significantly from design norms, plan a deeper service visit. That service is outside TM44 scope, but the recommendation belongs in the report with supporting observations.
Quantifying benefits without guesswork
TM44 requires cost-effective recommendations, which implies quantification. In industrial settings, energy estimates can be credible without being overly precise. The key is to connect measures with observable data.
For example, if head pressure runs at 36 bar when 30 bar would suffice under current ambient, estimate compressor power penalty using manufacturer data or generic performance ratios. If the measure is coil cleaning and control retuning, present a savings range rather than a single figure, and note seasonal variation.
For scheduling changes, use trends to quantify hours of unnecessary operation. If an AHU runs 168 hours per week at 7 kW fan power, and you can shut it down for 40 hours without affecting production, that is roughly 280 kWh per week saved on fans alone, plus associated cooling reduction. Numbers like these make it easier for plant managers to approve changes.
Edge cases and judgement calls
Not every system should be pushed to maximum efficiency. In pharma and food-grade environments, product integrity trumps energy every time. Some spaces require tight temperature and humidity bands to comply with validated procedures. The best TM44 reports make clear where flexibility exists and where it does not. If humidity can be loosened by 2 percent without affecting quality, say so and reference the controlling SOP. If it cannot, state that plainly and shift attention to alternatives like heat recovery or improved free cooling.
Another edge case is redundancy. N+1 systems often idle a backup unit. Regular rotation is needed to keep seals lubricated and detect faults. Rotation increases starts and might slightly dent efficiency, but it protects uptime. The assessor should acknowledge these trade-offs and suggest rotation schedules and run-hour balancing that minimize penalties.
Temporary cooling is a final edge case. Portable DX units or spot coolers multiply quietly, especially after a hot summer. Each unit adds uncontrolled load and often fights with central systems. A TM44 inspection should inventory these units and recommend either proper integration into the BMS or replacement with a permanent solution once the root cause is identified.
Turning findings into a practical action plan
The mark of a strong TM44 output is a set of recommendations that plant staff can carry forward without endless meetings. Group actions into three buckets: no-cost tuning, low-cost maintenance or minor hardware, and planned capital or controls projects. The time horizon matters. Suggest measures that can be executed within existing maintenance windows, and where a shutdown is needed, align with known outages.
For example, no-cost tuning might include updating AHU schedules, widening deadbands, enabling or fixing economiser logic, and removing stale overrides. Low-cost items often involve coil cleaning, belt replacements, damper repairs, sensor replacements, and adding differential pressure sensors across filters if not present. Planned projects might be VSD retrofits on fans and pumps, chiller sequencing upgrades, tower refurbishment, or partial plant recommissioning after major changes.
A concise one-page summary helps decision-makers. List the action, the expected energy saving range, the cost band, and any operational risk or requirement like scaffolding or hot works. Provide references to trend evidence or photos that support the case. Avoid generic language and tie each recommendation to a specific asset and observation.
A brief anecdote: the stuck economiser that wasn’t
On a spring visit to a packaging plant, the BMS showed the main AHU economiser modulating normally. Yet chilled water use stayed stubborn even on cool days. On inspection, the outside air damper shaft turned, the position feedback changed, and the actuator seemed fine. A small test revealed the issue: the damper blades had sheared off their internal couplings and freewheeled with the air pressure. The actuator moved the shaft, not the blades. A temporary bar clamp confirmed the diagnosis. Once repaired and recalibrated, the plant saw chilled water valve positions drop by 30 to 40 percent during shoulder months. It was not a complex fault, just a hidden one. TM44’s framework allowed it to be captured, quantified, and fixed.
Common pitfalls to avoid
The most frequent weakness in TM44 assessments is generic advice. “Consider VSDs,” “Consider free cooling,” and “Improve maintenance” mean little without context. Recommendations must include the system name, the observed behavior, and the specific change. Another pitfall is ignoring the BMS as it is. If trends are missing, say so and ask for them. If alarms have been suppressed, document that risk. Finally, do not overlook the economic reality of industrial schedules. A recommendation that requires full production stoppage with no workaround will gather dust. If a task needs a shutdown, propose interim steps, like staged coil cleaning during weekend half shifts, until a plant-wide outage occurs.
What good looks like
When TM44 best practices are applied, industrial sites typically realize immediate savings in the low single digits of total electricity within weeks, and more as maintenance and control upgrades roll in. The broader outcome is clarity. Plant teams know which assets drive their energy profile, how controls behave under real loads, and where future capital will yield the most. The inspection becomes a living document that shapes maintenance strategy and capital planning, not a report that vanishes into a shared drive.
Treat TM44 as a structured conversation between the system as designed, the system as operated, and the system as needed. That mindset shifts the assessment from compliance to performance, which is where industrial HVAC earns its keep.