Process Cooling in a Michigan Plant: Chillers, Cooling Towers, and What It Costs You When It Goes Down

Process cooling is a different job than comfort cooling

Comfort cooling has a forgiving target. If the office drifts two degrees on a hot afternoon, someone complains and it gets fixed the next morning. Process cooling has no such margin. An injection molding tool that runs warm changes cycle time and part dimensions. A spindle running above its coolant temperature loses accuracy and wears. A welder or a laser with inadequate cooling faults out and takes the cell down with it.

The other difference is the calendar. Comfort cooling is a summer system. Process cooling is a year-round system, because the heat you are removing comes from your equipment rather than the weather. A plant running two shifts in February is still rejecting the same process heat it rejects in July. That single fact drives most of what follows.

Where process cooling shows up in Southeast Michigan plants

Our region's manufacturing base means we see a fairly consistent set of applications. Injection molding and blow molding, where mold temperature control is the whole ballgame. Machining, where spindle and coolant cooling protect both accuracy and tool life. Tool and die work, EDM, laser cutting, and press welding equipment. Hydraulic systems and air compressors that need their heat taken somewhere. Plating, anodizing, and finishing lines. Food production with product and room cooling requirements. And increasingly, server and control rooms that grew into something that needs real cooling rather than a window unit.

Each of those has a different tolerance for temperature swing and a different consequence when it drifts, which is why we start by asking what the process actually needs rather than what tonnage someone wrote on a spec sheet. Stability often matters more than raw capacity. A system that holds a tight band at moderate load will serve a molding operation better than a larger one that cycles.

Air-cooled chillers

An air-cooled chiller rejects heat directly to outdoor air through a condenser coil, usually from a packaged unit sitting outside or on the roof. There is no cooling tower, no condenser water loop, and no water treatment program to run.

That simplicity is the appeal. Installation cost is lower, the mechanical room footprint is smaller, maintenance is more straightforward, and there is less to fail. For small and mid-size loads, and for plants without staff to manage a water treatment program, an air-cooled machine is frequently the right call.

The tradeoffs are efficiency and ambient sensitivity. Air-cooled units lose capacity as outdoor temperature rises, which is exactly when your building is working hardest, so sizing has to account for design-day conditions rather than average ones. They are noisier, they need clear airflow around them, and in a dirty or lint-heavy environment the condenser coils need regular cleaning or performance falls off quickly.

Water-cooled chillers and cooling towers

A water-cooled chiller rejects heat into a condenser water loop, which carries it to a cooling tower where evaporation dumps it to the atmosphere. Because evaporative cooling can reach a lower temperature than the outdoor air itself, these systems generally run more efficiently than air-cooled equipment, and the advantage grows with load size and on the hottest days.

What comes with that is a water system to manage. Condenser water needs a treatment program to control scale, corrosion, and biological growth, and that includes responsible management of Legionella risk, which is a genuine health and liability matter rather than a box to check. Towers need cleaning, fill inspection, drift eliminator checks, basin maintenance, and makeup water. There are more pumps, more piping, and more components that can fail.

For a larger plant running significant load most of the year, the efficiency usually justifies all of that. For a smaller operation without maintenance staff, it often does not. We would rather talk through which situation you are in than default to one answer.

Free cooling: the part of our climate worth designing for

Here is the advantage of operating in Michigan. For a large share of the year, outdoor air is colder than the water temperature your process needs. When that is true, you can reject process heat without running compressors at all, using a dry cooler or a waterside economizer arrangement with a glycol loop. The compressors idle, and you are essentially paying for pumps and fans.

How much of the year this covers depends entirely on your required supply temperature. A process that can accept 60°F water gets far more free cooling hours in our climate than one that needs 45°F. That is why the first design question is what temperature the process actually requires, not what temperature the current system happens to produce. We have found plants running a setpoint nobody could justify, set years earlier by someone who has since retired, and simply raising it where the process allowed opened up months of free cooling.

We do not put a savings percentage on this before running your numbers, because it depends on your load profile, your setpoint, and your run hours. What we can say is that in a climate with our heating season, free cooling is worth evaluating on any process cooling project rather than treating it as an exotic add-on.

Winter is also where process cooling fails

Everything outdoors that holds water is a freeze risk from November through March. Cooling tower basins, condenser water piping, dry coolers, makeup lines, and any piping routed through an unheated space or along an exterior wall. The failures we get called for are rarely subtle: a split line, a flooded mechanical room, and a production line down.

Glycol is the usual protection, but glycol is not a set-and-forget solution. Concentration drifts as water is added over time, and a loop that tested fine three winters ago may no longer be protected to the temperature you think it is. Annual glycol testing before heating season is one of the cheapest insurance policies in a plant. Basin heaters, heat trace, and freeze-protection controls all need to be verified as working, not merely present.

Holiday and summer shutdowns deserve particular attention. A building that sits idle over a long weekend in January, with the heat setback and nobody walking through, is where most freeze damage happens. A short written shutdown checklist covering what gets drained, what stays running, and who checks the building prevents nearly all of it.

What downtime actually costs, and what redundancy costs

The honest way to size redundancy is to price an hour of stopped production and compare it against the equipment. Plenty of plants discover that a single hour of downtime exceeds what a backup pump would have cost, and a full day exceeds the price of real capacity redundancy.

There is a middle ground between a single chiller and full N+1 capacity. Two smaller chillers instead of one large one give you partial production during a failure and better part-load efficiency the rest of the time. Spare pumps, spare motors, and critical valves kept on the shelf turn a multi-day parts wait into a same-day repair. And pre-plumbed connection points for a rental chiller are inexpensive to install during a project and invaluable at two in the morning in July, when the rental company can be connected in hours instead of a day of improvised piping.

That last one is worth pressing on. Rental chillers are available, but the hookup is what costs you the time. Deciding where a rental would connect while a system is being installed is far cheaper than figuring it out during an outage.

The maintenance that prevents the emergency call

Process cooling responds well to preventive maintenance because most failures announce themselves in the data long before they stop the line. Condenser approach temperature creeping up means the coil or the tower fill is fouling. A shrinking delta-T across the evaporator points at flow problems or a plugged strainer. Rising compressor amp draw, falling refrigerant charge, and vibration changes on pumps are all visible early if someone is recording them.

A workable program looks like this: scheduled visits with logged readings rather than a checkbox, condenser and tower cleaning on a frequency that matches how dirty your environment actually is, strainer service, a water treatment program with documented results for water-cooled systems, annual glycol testing, refrigerant and leak checks, control and safety verification, and a review of the trend data with you rather than a service ticket left on a desk.

The trend data is the part most often skipped and the part that pays. A chiller that is losing half a degree of approach temperature every month is telling you when it will fail, and that is information you can schedule around.

Doing the work without stopping production

Process cooling projects touch mechanical, electrical, piping, controls, and sometimes structural work, and they land in a building that cannot simply close for a week. We plan around your production calendar: phased tie-ins, temporary cooling where it is warranted, work scheduled into shutdown windows and off-shifts, and permits and inspections sequenced at the start rather than discovered halfway through.

The best time to do this work is when you are not desperate. A chiller replaced on your schedule in March costs less, disrupts less, and gets commissioned properly. The same chiller replaced in July after a failure costs more in every direction. If yours is aging, the useful move is to get it assessed while you still have options, the same way we approach heating a warehouse or shop, where planning ahead is most of the savings.