Most AHU specifications are built from design assumptions. Occupancy estimates, standard diversity factors, climate data from reference years, and equipment heat gains drawn from drawings that may never have reflected how the building actually operates.
When those assumptions were reasonable, the resulting system performs adequately. When they weren't — and in most commercial buildings, they weren't entirely — the new AHU inherits the same mismatches as the one it replaces.
There is a better starting point. Buildings with an operating BMS already hold months or years of evidence about how the space actually behaves: how hard the system works, when it struggles, where loads concentrate, and how demand shifts across the day and the season. That data exists before a specification is written. The question is whether it gets used.
This article is a practical guide to extracting the right signals from BMS trend logs and translating them into AHU specification inputs — coil selection, fan sizing, filtration staging, and control requirements — grounded in measured behaviour rather than projected averages.
Load calculations produce a snapshot. They estimate peak demand under defined conditions and size equipment to meet it. That process is necessary, but it captures the building at one theoretical moment, not across the variability of real occupancy, seasonal shift, and operational change.
Diversity factors — the assumptions that govern how much of the theoretical peak load actually coincides — are among the weakest inputs in any load model. Standard diversity figures are derived from building type averages. They do not reflect the specific usage patterns of a particular floor, tenant, or operating schedule.
When a building has been running for two or more years, there is no need to estimate diversity. The BMS has measured it. Fan speed trends show whether peak airflow demand is reached regularly or rarely. Runtime profiles show how long the system operates above 80% capacity versus how much time it spends at part-load. Zone temperature deviation patterns show where heat concentrates and where loads are consistently lower than assumed.
Specifying a replacement AHU from this evidence base produces a system calibrated to the building as it is, not as it was assumed to be.
Not all BMS data carries equal weight for specification purposes. The following five signal categories are the most directly translatable into equipment selection parameters. Each one maps to a specific decision in the AHU configuration.
Read: Faulty HVAC Sensor Data and System Performance
Fan speed logs over a 12-month period reveal the actual demand profile the unit operates against. The key metric is not peak speed — it is the distribution of operating hours across the speed range. A system that spends the majority of its runtime between 45% and 65% fan speed is telling you that turndown performance matters more than peak capacity. Fan selection for a replacement AHU should prioritise stable, efficient operation across that observed range, not simply the ability to reach maximum airflow. If the existing unit spent sustained periods at or near maximum speed, that indicates either undersizing or load growth that must be accounted for explicitly in the new specification.
Read: Smart Systems Reduce HVAC Downtime
The temperature difference between supply and return air reflects how much thermal work the coil is doing under real conditions. Consistent, stable Delta-T across varying load periods indicates a well-matched coil. A declining Delta-T trend over time, without corresponding changes in setpoint, typically points to coil fouling or valve degradation. More importantly for specification: if the measured Delta-T across the duty cycle is consistently lower than the design value, it may indicate that the coil was oversized for sensible load but undersized for latent demand, or that supply temperatures were being compensated by control rather than delivered by coil performance. Replacement coil selection should be benchmarked against the observed Delta-T range, not the design-stage figure.
Read: Smart Sensors for High-Efficiency Air Conditioning
The rate at which filter pressure rises between service intervals is a direct measure of particulate load in the airstream. In South African commercial buildings, this varies substantially by location, building type, and ventilation strategy. A filter that reaches its end-of-life pressure drop in six weeks rather than the assumed twelve is telling you that the filtration staging in the replacement AHU needs to be reconfigured — typically by adding a pre-filter stage to protect the main filtration bank and extend service intervals. It also sets the static pressure envelope the fan must operate against as filters load, which is a direct input to fan selection.
Zone-level temperature logs reveal where the existing system consistently over- or under-delivers. Zones that repeatedly drift above setpoint during occupied hours despite sufficient nominal capacity indicate either a distribution problem or localised load growth. Zones that remain consistently below setpoint point to oversupply or poor zoning. Both patterns inform the supply air volume and zoning strategy for the replacement AHU. Where a zone has developed a persistent heat load not present in the original design — dense IT equipment, reconfigured open plan to enclosed offices, or extended operating hours — that load should be explicitly included in the replacement specification rather than absorbed into a generic diversity allowance.
Normalising energy consumption data against heating and cooling degree-days produces a measure of system efficiency that is independent of weather variation. A rising energy intensity trend over two or more years, under similar occupancy and operational conditions, indicates declining system efficiency — most commonly from coil fouling, control drift, or equipment operating outside its optimal range. This trend establishes a performance baseline against which the replacement AHU can be specified and subsequently benchmarked. It also provides defensible evidence for the selection of higher-efficiency components where the lifecycle cost justification needs to be made to building owners or project committees.
Raw BMS data requires interpretation before it becomes a specification input. The translation steps below address the most common signal-to-specification relationships.
From fan speed distribution to fan selection criteria: If trend logs show that the system operates below 60% of design airflow for more than 60% of its runtime, the fan selection criteria should shift toward optimised part-load efficiency rather than peak capacity. Electronically commutated (EC) fans or variable speed drives with carefully selected motor-fan combinations provide stable, efficient operation across the observed range. Specifying for peak demand alone in this scenario produces a fan that operates inefficiently throughout most of its service life.
From Delta-T behaviour to coil configuration: Sustained low Delta-T values at moderate load conditions — where the coil should be performing well — suggest face velocity or coil depth mismatches in the existing unit. The replacement specification should define coil face velocity limits and minimum coil row depth based on the observed operating range, not solely on peak sensible load. Where measured latent loads are higher than the design model predicted, coil fin spacing and surface geometry should reflect moisture removal requirements at typical operating conditions, not just peak sensible performance.
From filter pressure trends to filtration staging: The measured rate of pressure rise across filter stages, combined with the total static pressure budget available to the fan, determines how the filtration section should be staged. A two-stage system where the first stage loads quickly compresses the fan's available static pressure margin and forces higher speeds to maintain airflow — visible as a rising fan speed trend correlating with filter age. A replacement AHU specification that adds a coarser pre-filter stage before the main bank extends the useful life of the primary filter, flattens the pressure rise curve, and reduces the fan energy penalty across the service interval.
From zone deviation patterns to supply air volume and zoning: Persistent hot zones that do not respond to control adjustment are a specification input, not a commissioning problem to solve later. Where BMS data confirms that specific zones have developed loads not present in the original design, the replacement AHU specification should address supply air volume allocation and, where necessary, zoning configuration. Attempting to solve an undersupply condition through control tuning of an identically-sized replacement unit will reproduce the same deviation pattern.
Operational data establishes how a system behaves under existing conditions. It does not automatically account for planned changes that will alter those conditions after the replacement AHU is installed.
Tenant changes, floor reconfigurations, increased IT density, extended operating hours, or planned occupancy increases are not captured in historical BMS data. These must be identified separately and added to the specification as adjustment factors. The BMS baseline establishes the minimum performance requirement; planned changes establish the margin above it.
Similarly, BMS data reflects the performance of the existing system, including its limitations. A zone that has never been adequately cooled will show persistent deviation — but that deviation reflects an existing deficiency, not the correct target load. Where the existing system is known to be undersized for a zone, the BMS data requires correction before it becomes a specification input.
The practical output of this process is a specification document where each key parameter — design airflow, coil selection criteria, fan performance envelope, filtration staging, static pressure allowance, and control requirements — is accompanied by the BMS data that supports it.
This is not a standard approach. Most AHU specifications are produced from load calculations alone, with BMS data reviewed separately if at all. Combining both produces a specification with two layers of support: the theoretical load model and the measured operating evidence. Where they agree, confidence in the selection is high. Where they diverge, the divergence itself is a specification input — it indicates either load growth, model error, or a system deficiency that the replacement must address.
For consulting engineers and project managers specifying replacement or upgrade AHUs in buildings with operating BMS infrastructure, this approach reduces the risk of reproducing the performance gaps that led to the replacement decision in the first place.
Air Options designs and manufactures custom air handling units for commercial, medical, data centre, and specialised environments. Where buildings have operational BMS data available, our engineering team works with consulting engineers and project teams to incorporate measured operating behaviour into AHU configuration — coil selection, fan specification, filtration staging, and control compatibility — alongside standard load calculations.
If your project involves replacing or upgrading an AHU in a building with existing operational data, contact Air Options to discuss how that data can be applied to the specification process, or request a quote here.
Get the latest updates in your email box automatically.
Your nickname:
Email address:
Subscribe
Request A Quote