Cooling System Inspection Checklist
A cooling system inspection checklist template for the system as a whole rather than for any one machine in it, built around delta-T measured at every stage and compared with design. A chilled water system running a low temperature difference moves far more water than it needs to, every hour it runs, and nobody reports it because everything still works. It shows up in the pumping bill, and only if somebody reads that properly.
- Delta-T measured at nine stages, each against design
- Ten causes of a low delta-T, worked before buying plant
- Overrides and hand positions checked and dated
- Load and conditions recorded at the time of the survey
Cooling System Inspection Checklist
Delta-T, stage by stage
| # | DELTA-T ACROSS THE SYSTEM | FLOW | RETURN | DESIGN |
|---|---|---|---|---|
| 1 | Chiller evaporator | |||
| 2 | Primary circuit | |||
| 3 | Secondary circuit |
The document you will get. Download for the full, editable file.
Who this cooling system inspection is for
One system, four people, and only one of them is looking at a machine.
Whoever carries out the survey
A sequence that starts at the chiller and ends at a grille, with delta-T at every stage in between. The reading matters more than the tick on every line of it.
The energy or sustainability lead
Delta-T shortfall and the pumping cost behind it. It is the one line on a maintenance document that converts directly into consumption with no further work.
The facilities manager
Zones not achieving setpoint, mapped against where the complaints came from. Two zones off setpoint and a system-wide hydraulic fault are the same finding seen from different ends.
Whoever is being asked to fund new plant
The capacity question, answered last. It is the field that stops a hydraulic problem being solved with a bigger machine.
Where whole-system cooling surveys get commissioned
The structure suits any central cooling system with a distribution circuit. What changes is what the building does when the cooling underperforms. If you run one of these, the sector page goes further than the template does.
- Commercial real estate
Multi-tenant buildings on central chilled water, where pumping is a landlord cost and poor delta-T is recovered from occupiers who did not cause it.
Portfolio maintenance software - Corporate facilities
Office cooling where the complaint pattern and the energy data describe the same hydraulic fault from two different ends.
Corporate facilities software - Healthcare
Clinical cooling, where a zone off setpoint is a clinical issue and the survey cannot wait for a convenient shutdown.
Healthcare maintenance software - Education
Campus systems serving several buildings from one plant room, where the distribution is long and nobody has measured it since handover.
Campus maintenance software - Retail and malls
Shopping centers and large retail, where terminals are obstructed by stock and the system is blamed for it.
Retail maintenance software - FM service providers
Contracted plant maintenance, where the machines are serviced well and the system between them belongs to nobody.
FM service provider software
What a cooling system survey should cover
A cooling system inspection is a survey of a whole cooling system rather than a service of one machine. It measures delta-T at every stage against design, works through the causes of a low figure, checks distribution and hydraulics, the terminals, and controls and overrides, then puts an impact and a priority against every finding.
A. Measure first, diagnose second, cost it third
| Field | What goes in it | Why it earns its place |
|---|---|---|
| System details | Type, design capacity, design flow and return, design delta-T, and what it serves | The design delta-T is the figure everything else is measured against. Without it the survey records temperatures and proves nothing. |
| Conditions and load at the time | Recorded explicitly, alongside the reason for the inspection | A survey carried out on a cool morning at low load proves almost nothing, and the reader has no way to know that unless this field is filled in. |
| Delta-T across the system | Nine stages, each with flow, return, measured delta-T, design and variance | Stage by stage, because the point at which the figure collapses tells you where the problem is. A good delta-T at the chiller and a poor one at the terminals is a distribution fault, not a plant fault. |
| What a low delta-T usually means | Ten specific causes, each marked OK, Issue or not applicable | Three-port valves where two-port were designed, valves passing, coils fouled on either side, bypasses open, pumps at fixed speed, secondary flow exceeding primary, and a setpoint lower than it needs to be. |
| Distribution and hydraulics | Twelve checks from pump curves to the vapor barrier | Includes balancing valves present, set and recorded, which is the one that is most often missing entirely rather than wrong. |
| Insulation and vapor barrier | Continuous, including at supports and valves | Missing insulation on chilled pipework wastes cooling and condenses water onto ceilings at the same time. The second shows up as a fabric problem long before anybody blames the cooling system. |
| Terminals and served spaces | Nine checks, including zones in simultaneous heating and cooling | Local heaters and desk fans are on this list deliberately. They are the clearest evidence a zone has given up on the system, and they never appear in a plant room. |
| Controls and operation | Nine checks covering schedules, sequencing, free cooling, trends and alarms | Setpoints and schedules measured against how the building is actually used, rather than against how it was used when the system was commissioned. |
| Manual overrides | In place, and for how long | A valve put into hand during a fault six months ago is still in hand. It will not show on the graphics and it will be plainly visible in the energy data, which is why both get checked. |
| Findings and recommendations | Area, finding, impact, recommendation, priority and an estimated cost | Impact is the field that makes the difference. A survey that lists faults without saying what each one costs gets read once and filed. |
| System position | Delta-T shortfall, zones off setpoint, findings by priority and the total estimate | A short block that can be read on its own, because the person who decides whether to fund the work will not read the whole survey. |
| Is capacity adequate | Answered last, and only after the hydraulics are understood | The most consequential field on the document, and the reason it sits at the end rather than the beginning. |
Delta-T earns the top of this document because it is the one measurement that describes the system rather than any part of it. A chilled water system is sized on the assumption that water leaves the plant cold, picks up a specified amount of heat on its way round, and comes back warmer by a known number of degrees. That difference is what the pumps, the pipe sizes and the terminal coils were all selected for. When the real difference is smaller than the design figure, nothing breaks and nothing alarms, because the building still gets cool air. What happens instead is that the system has to circulate more water to move the same heat, and the arithmetic there is unforgiving: a system achieving 3.9 against a design of 6.0 is moving roughly one and a half times the intended flow, and because pump power rises with something close to the cube of flow, the pumping energy is two to three times what it should be, continuously, for as long as the system runs. That is a large recurring cost with no symptom attached to it, which is exactly why it survives for years in buildings that are otherwise well maintained. The second reason to lead with it is diagnostic. Measured at one point, delta-T is a number. Measured at nine points from the evaporator out to the zone terminals, it localizes the fault: a healthy figure at the chiller that degrades through the secondary circuit says the problem is in the distribution, while a poor figure everywhere including at the plant says something upstream is wrong. That is why the template asks for a stage-by-stage table rather than a single system figure, and why the list of ten causes sits directly underneath it rather than in an appendix.
B. What it looks like filled in
One secondary circuit, measured at full load on a warm afternoon. Nothing has failed and the finding is expensive.
| Line | Value |
|---|---|
| Stage | Secondary circuit |
| Flow and return | 6.2 and 10.1 |
| Delta-T achieved | 3.9 against a design of 6.0 |
| Shortfall | 2.1 |
| Finding, priority | Pumping cost and capacity, High |
| Recommendation | Survey three-port valves on zones 2 and 3 |
A return temperature of 10.1 against a flow of 6.2 is a delta-T of 3.9, and the system was designed for 6.0. As a temperature it is unremarkable. As a ratio it is the most expensive line in the survey, because the system is circulating roughly half as much water again as it was meant to in order to move the same heat, and pump power does not rise in step with flow but with something close to its cube. That is the two to three times figure, and it applies every hour the system runs rather than at a peak. The recommendation is deliberately narrow: survey the three-port valves on two specific zones, rather than a general instruction to investigate. Three-port valves fitted where two-port were designed keep water moving through a bypass whether the coil needs it or not, which blends cool return water back into the circuit and holds the return temperature down. It is the commonest single cause on the list and it is a hydraulic fault, not a capacity one. The machines underneath this survey have their own documents and their own readings: the chiller maintenance checklist for the machine itself, and the cooling tower maintenance checklist for the loop that rejects the heat.
Word for the version a surveyor completes on site, Excel for the one that calculates delta-T and variance at every stage and holds a Survey Log so the figure can be tracked across inspections, and PDF for the report that goes to whoever is being asked to fund the work. Free, and yours to rebrand.
How do you inspect a chilled water system?
Measure at load, work down from the plant, and leave the capacity question until last. Six steps.
Record the design figures and the load
Design flow, return and delta-T, then the actual load and conditions at the time of the survey. Readings taken on a cool morning describe a cool morning and nothing else.
Measure delta-T at every stage
Evaporator, primary, secondary, riser, each zone of terminals, condenser and free cooling. Stage by stage, because where the figure collapses is where the fault is.
Work the ten causes before proposing plant
Three-port valves, passing valves, fouled coils, open bypasses, fixed-speed pumps, secondary flow over primary. A system with the right plant and the wrong hydraulics looks exactly like one that is undersized.
Check the hydraulics and the insulation
Pump differential against the curve, balancing valves set and recorded, strainers, air vents, and the vapor barrier on chilled pipework where a gap condenses water onto a ceiling.
Look for overrides and hand positions
And write down how long each has been there. An override left on during a fault six months ago will not appear on the graphics and will be obvious in the trend data.
Put an impact against every finding, then answer capacity
Cost or consequence on each line, not just a description. Then, last, decide whether the system is genuinely short of capacity.
A system survey versus a machine service
Two documents that both look at cooling and answer questions that do not overlap. Most buildings buy the second and never commission the first.
| Aspect | The system survey | The machine service |
|---|---|---|
| What it looks at | Everything between plant and grille | One machine |
| Its headline measure | Delta-T against design | Approach, pressures, efficiency |
| What it finds | Hydraulic and control faults | Wear, fouling, failing parts |
| Who it is for | Whoever pays for energy | Whoever keeps the plant running |
| How often | Periodically, or on a complaint | On a service cycle |
| What it cannot see | A machine about to fail | A system wasting energy |
The gap between these two documents is where the expensive problems live, because a system can be made entirely of well-serviced machines and still perform badly. Every machine in a chilled water system is somebody's contracted responsibility. The chiller is serviced, the tower is dosed and cleaned, the air handling units have their filters changed, and each of those visits produces a satisfactory report. Nothing in any of them looks at the pipework, the valves, the balancing or the control strategy that connects them, because none of that is a machine and so none of it is on anybody's service schedule. That is how a building ends up with a plant room full of healthy equipment and a system moving half as much water again as it needs to. It is also the mechanism behind the single most costly mistake in this area: a building that cannot hold temperature in a hot spell concludes it is short of cooling, specifies a larger chiller, installs it, and discovers the following summer that the new machine runs at the same poor delta-T as the old one, because the constraint was never capacity. The survey costs a fraction of the plant and should always come first. For the machines underneath it, use the chiller maintenance checklist and the cooling tower maintenance checklist, and for the air side of the same building the ventilation inspection checklist and the HVAC maintenance checklist.
When the template starts to feel limiting
The file holds one survey well. Four things get difficult the moment the survey is finished and filed.
Delta-T is a snapshot, once
The figure that matters is whether remedial work actually moved it, and that needs two surveys side by side rather than one in a folder.
Findings have no owner once the report is issued
A priority and an estimated cost are on the page, and nothing connects either to a job that somebody is accountable for finishing.
Overrides go back on
The survey catches the ones in place on the day. Nothing watches for the next valve put into hand during the next fault.
The energy data lives somewhere else
Delta-T is a maintenance reading and pumping cost is a utility report, and the argument only works when the two are read together.
What running this in Facilio looks like
The survey does not change. What changes is that delta-T becomes a trend and a finding becomes a job.
Asset Intelligence
Readings attach to the system
Delta-T at each stage accumulates against the plant, so the question of whether the remedial work fixed the hydraulics has an answer rather than an opinion.
Work Order Intelligence
Findings become owned work
A High priority finding with a cost estimate turns into a job with a target date, instead of a line in a report somebody has to re-read.
Ops Performance Intelligence
Overrides and drift surface
A hand position left on, or a delta-T degrading across surveys, is visible without waiting for the next annual inspection.
Audit Report Intelligence
The survey stays with the system
Design figures, measured figures and what was done about them sit together, which is what makes the capacity question answerable later.
Hallucination-free by design. Atom AI answers from the records in your tenant rather than generating plausible text, so an empty field reads as empty rather than filled in for you.
Frequently asked questions
What is a cooling system inspection checklist?
It is a survey of a whole cooling system rather than a service of one machine in it. This one measures delta-T at nine stages against design, then works through ten causes of a low figure, twelve hydraulic checks, nine at the terminals and nine on controls.
The individual machines have their own documents. This is the layer above them, covering everything between the plant and the grille.
What is delta-T and why does it matter?
It is the temperature difference between the water leaving the plant and the water coming back, and it is what the pumps, pipe sizes and coils were all selected for.
When it runs below design the system circulates more water to move the same heat. A system at 3.9 against a design of 6.0 moves roughly half as much water again, and because pump power rises with close to the cube of flow, that is two to three times the pumping energy.
Why does nobody notice a low delta-T?
Because the building is still cool. Nothing fails, nothing alarms and no occupant complains, so there is no symptom to report and nothing to raise a job against.
It is a cost rather than a fault, and it only becomes visible in a pumping bill that somebody reads carefully, which is why it survives for years.
What causes a low delta-T?
Most often three-port valves fitted where two-port were designed, control valves passing or stuck open, coils fouled on the air or water side, bypass lines open, balancing valves set wrong, and pumps running at a fixed speed.
Also secondary flow exceeding primary flow, a chilled water setpoint lower than it needs to be, oversized terminals, and blocked strainers restricting flow through coils.
Should we fit a bigger chiller?
Not until the hydraulics have been surveyed. A system with the right plant and the wrong hydraulics behaves exactly like one with too little capacity, and the two are indistinguishable from inside the building.
A great many buildings have bought a bigger chiller to solve a hydraulic problem and found the new machine running at the same poor delta-T as the old one. The capacity question belongs at the end of this document for that reason.
When should the readings be taken?
When the system is under real load, and the template asks for the load and external conditions to be written down alongside the readings.
A survey carried out on a cool morning at part load proves very little, and a reader has no way to discount it unless the conditions are on the page.
Why check for manual overrides?
Because a valve or a damper put into hand during a fault is almost never put back. Six months later it is still in hand, holding a circuit open or a setpoint fixed.
It is invisible on the control graphics, which show the commanded position rather than the actual one, and it is plainly visible in the energy data if anybody compares the two.
Does this replace servicing the chiller?
No. This is a survey of the system; the chiller, the cooling tower and the air handling units each have their own service documents and their own readings.
The two are complementary and the gap between them is where the expensive problems live, because a system can be made entirely of well-serviced machines and still waste a great deal of energy.
In one paragraph
A cooling system inspection works when delta-T is measured at every stage and compared with design, because a system running below its design difference moves far more water than it needs to and nothing about it looks broken. Record the load and conditions at the time, work the ten causes of a low figure before anybody proposes new plant, check for overrides left in hand during old faults, put an impact against every finding, and answer the capacity question last.
Nothing is broken and it is costing you
Measure at load, stage by stage, diagnose before you specify, and leave the capacity question until the hydraulics are understood. Once delta-T is a figure in a folder rather than a trend, or a finding has a cost but no owner, the paper version has reached its limit.
