September 2026
Two constraints, and everything follows from them
A data hall cannot take water and cannot take downtime. Every cleaning conversation in a critical facility starts and ends with those two sentences, and most methods fail one of them immediately.
Which is why the electrical and mechanical plant in these buildings tends to go uncleaned for years. Not because nobody noticed, but because the obvious options were all unacceptable and there was no third one on the list.
What the particulate actually does
The loading in a critical facility is fine rather than heavy. Nobody has grease on their switchgear. What they have is a settled layer of fine dust, and it causes trouble in three separate ways that get diagnosed separately.
Thermal first. Dust on a heat sink, a fin stack, a fan blade or a motor frame is insulation. The equipment cannot shed heat at its designed rate, so it runs hotter for the same load. Hotter means shorter component life and more energy for the same work, and it never presents as a cleaning problem.
Airflow second. Loaded fan wheels and blocked filters move less air. That shows up as a cooling capacity that quietly is not what the drawings say, and on a mechanical plant that has been in service a decade it is often significant.
Electrical third, and this is where humidity matters here. A dust layer on an insulator, a bus bar or a standoff is inert while it is dry. In South Florida it does not stay dry. It cycles moisture out of the air every night, and a damp conductive layer is a leakage path across the gap a designer put there deliberately.
Why the familiar options are out
Water is not a conversation. Nobody is washing gear in a data hall.
Compressed air is the one people actually reach for, and it is the worst choice available. It does not remove particulate, it relocates it. Dust blown off a switchgear face is now airborne in a room designed around controlling exactly that, and it settles on equipment downstream, including inside cabinets you cannot easily open. On a live system it is a hazard on top of being ineffective.
Vacuums and wipes work on surfaces a person can physically reach, which in a motor control center or a busway run is a small fraction of what needs attention.
Chemicals bring residue, fumes and a material compatibility question on every plastic and label in the cabinet.
Abrasive is not in the discussion for reasons that need no explanation.
What makes a dry thermal method fit
Four properties, each mapping to one of the constraints above.
The media is non-conductive, and it leaves as gas rather than staying in the cabinet.
The process is low-moisture, so nothing is left standing in an enclosure, a winding, a harness or a terminal block. That is the technical reason this method is used on electrical equipment at all.
Contamination is captured rather than redistributed, which is the opposite of what a blow-out does. Containment planning is most of the setup time on this work, and that is as it should be in a controlled environment.
And the work happens in place, which is what makes it possible to finish inside the window you were given rather than negotiating for a longer one.
The equipment list runs through switchgear, busway and bus duct, UPS equipment, standby generators, cooling and mechanical plant, and fans and air handling.
The window is the real negotiation
On every job in a critical facility the technical question is easy and the scheduling question is the whole thing.
What we need before we agree to anything: the real window length including handover time, which equipment is coming out of service and under whose procedure, what containment the facility requires, what the escort and access rules are, and who else is in the room during that period. Most of that is quicker to send through the contact page than to cover on a call.
The honest answer we give more often than customers want to hear: scope one system properly rather than four partially. A room where the switchgear is finished and the mechanical plant is untouched is a better outcome than four systems at seventy percent, because the second version has to be redone.
The first clean on equipment that has not been touched in years takes considerably longer than every clean afterward. Knowing that changes how you plan the first visit, and it is the reason a trial on one line-up beats putting this on the critical path of a major outage sight unseen.
Where the particulate is coming from
Worth diagnosing before you schedule cleaning, because if the source is still active you will be back.
Construction is the biggest single contributor and the most common. Any build-out, core drilling, cutting or fit-out in or near the space puts fine debris into the air, and it settles into every electrical room on the floor whether the work was in that room or not. A facility that has expanded twice has been dusted twice.
Outside air is second. Down here that means humidity, and within a few miles of the coast it also means a fine salt loading through economizer and makeup air paths. Salt is worse than ordinary dust because it holds moisture, which is what turns a cosmetic layer into a leakage path.
Third is the building itself: concrete dust off unsealed floors and slab edges, degrading filter media shedding fibers, and cable pulling in trays overhead.
Practical consequence: if a build-out is scheduled, clean afterward rather than before. And if you have the choice, get the cleaning done in the same outage as the electrical testing, because the gear is already isolated and the incremental downtime is zero.
Carbon dioxide in a contained room
This has to be said plainly because a critical facility team will ask, and should.
Carbon dioxide is heavier than air and displaces oxygen. In a large open hall that is a non-issue. In an electrical room, a vault, a plenum or any small contained space, mechanical ventilation and monitoring are requirements rather than precautions, and they get planned as part of the job.
Anyone who quotes work in a confined space without raising this is telling you something about how they operate, and a critical facility is not the place to find that out on the day.
Questions from critical facilities teams
Can you clean anything energized?
No. Equipment is de-energized under your own electrical safe work procedure by your own people. We plan around your isolation rather than asking you to work around us.
What about particulate migration into adjacent spaces?
That is the question that shapes the containment plan, and it is the reason we want to walk the space before the window rather than on the morning of it.
Will it affect labels, insulation or plastics?
Settings do the work, and the test area establishes them. Older brittle plastics and adhesive labels either get a light touch or get masked.
Do you work in colocation and mixed-tenant space?
Yes, and the access and notification rules there are usually the binding constraint rather than anything technical. Tell us the ruleset and we plan inside it.
Before the next planned outage
If you have a line-up or a mechanical room that has not been properly cleaned in years, the useful first step is a walk and a real duration for one system. Call Giancarlo at 786-302-6553 or use the contact page. We are mobile out of Miami across the South Florida counties, and we will tell you honestly what fits in the time you have.