HVAC & geothermal handbook

Name the requirement before selecting a concentration

Dow distinguishes freeze and burst protection in its DOWFROST product information. Those are different protection objectives, not interchangeable labels. The published values belong to that named formulation.

In a project brief, record which objective the equipment owner requires. Keep the lowest operating temperature separate from a shutdown or storage exposure. An entry that merely says “winter protection” leaves the acceptance decision unresolved.

Request a property envelope, not a single headline number

Our proposed purchasing check is to ask for the exact formulation, concentration basis, reference temperature and property-data revision together. Ask the designer to identify which temperature-dependent properties are required for their hydraulic and thermal evaluation.

Do not convert a manufacturer’s freeze-protection statement into a claim that the existing pump or heat exchanger will perform acceptably. That conclusion needs the actual system calculation and the applicable equipment requirements.

• Which conditions govern startup and continued circulation?
• Which data are measured, typical, or guaranteed specification limits?
• Who approves a formulation change and the corresponding design review?

Requirement → evidence → decision boundary

An editorial checklist for your review—not a table of product specifications.

Start with what is already in the loop

A purchase request should distinguish initial fill, routine replenishment, and a proposed change of formulation. We recommend recording the existing product identity and the evidence behind that identification. An old drum label is useful context, but it does not by itself document everything added during years of service.

When the existing fill is unknown, record it as unknown. Ask the equipment owner and fluid supplier to define the investigation and change-control requirements. Do not let a blank field become an implied compatibility approval.

Separate a condition report from an acceptance decision

For any service report, capture the sampling date, location, sample identifier, methods and the party interpreting the results. Ask that party which tests address the actual service concern and what limits apply to the exact formulation. This worksheet supplies a record structure, not universal test limits or a maintenance interval.

Dow describes DOWFROST as an inhibited formulation. That illustrates why a base-chemical name alone is an incomplete record of the product being purchased; it does not establish the condition of an installed fluid.

Close the receiving loop

Specify the container size and transfer connection alongside lot traceability, seal checks and the receiving hold point. Identify who can release the delivered material for use, and how a discrepancy is resolved. Keep handling requirements with the exact product SDS and the site procedure.

Requirement → evidence → decision boundary

An editorial checklist for your review—not a table of product specifications.

What is new in the source collection

DOE’s geothermal publications directory lists a heat-pump case-study collection introduced in 2024 and expanded in late 2025. It spans installations in different climate zones and system configurations. The directory is a useful starting point for finding projects with relevant conditions.

This brief reviews the directory’s scope, not the measurements in every underlying study. No energy-saving percentage or Alliance product-performance claim is inferred from it.

Build a transferability record

Our suggested next step is to select one relevant project and record its system boundary, climate, loads, comparison baseline and measurement period. Note whether results are measured or modeled and which operating conditions differ from your site.

Then separate system-level evidence from fluid-level evidence. A successful installation can motivate investigation of a similar system; it does not qualify a different fluid, concentration, ground-loop design or maintenance practice.

• Which project conditions resemble the proposed installation?
• Does the reported outcome include pumping and auxiliary loads?
• What fluid and operating data are actually disclosed?
• What remains a hypothesis requiring local engineering validation?

Requirement → evidence → decision boundary

An editorial checklist for your review—not a table of product specifications.

The base chemical is the smaller half of the product

Dow describes DOWFROST as 96.0 percent food grade propylene glycol plus food grade corrosion inhibitors, and DOWFROST HD as 94.0 percent propylene glycol with, in Dow’s words, “a specially designed package of industrial corrosion inhibitors.” Two products, one base chemical, two different inhibitor packages and two different glycol percentages. Nothing in the phrase “propylene glycol” distinguishes them, and nothing in it distinguishes either from a glycol supplied with no inhibitor at all.

That is the whole distinction. Uninhibited glycol is a chemical: it has a CAS number, an assay and a set of physical properties. An inhibited glycol is a formulation: a glycol, plus additives whose identity and concentration are usually the formulator’s, plus a claim about what that combination does in a metal system over time. The evidence you can obtain for the two is not the same kind of evidence, so the request you send a supplier should not be the same request.

What an inhibitor package is not

It is not a property of the glycol. Propylene glycol’s own records describe a stable material that, in the words of the Hazardous Substances Data Bank entry aggregated by PubChem, “at high temperatures, in the open, tends to oxidize, giving rise to products such as propionaldehyde, lactic acid, pyruvic acid, and acetic acid.” Those acids are part of what an inhibitor package is buffering against. A fluid that has consumed that buffer still contains glycol; it is no longer the product that was purchased.

It is not permanent either, and a federal design criterion says so plainly. UFC 3-410-01 §2-13.7 instructs designers to minimise the use of antifreeze in ground loops, to use propylene glycol where antifreeze is necessary, to keep concentrations to a minimum, and to “Coordinate glycol concentration with manufacturer to mitigate corrosion inhibitor degradation.” The criterion treats the inhibitor as something that degrades and as something coupled to concentration — which is why a purchase request that states only a freeze point has not stated the requirement.

And it is not transferable. A protection table, an inhibitor description or a service-life statement published for one named formulation is evidence about that formulation. A nominal grade or concentration does not establish interchangeability between products or between suppliers.

• Is the material being quoted a glycol, or a formulated inhibited fluid?
• If it is inhibited, which named formulation, and at what glycol percentage on which basis?
• Is it supplied ready to use or as a concentrate, and what dilution water does the manufacturer require?
• Who approves the choice, and does an existing fill constrain it?

Write the request so a quote can answer it

Our suggested request names four things together: the exact product or specification, the concentration with its basis and reference temperature, whether the material is a concentrate or ready to use, and the protection objective the equipment owner actually requires. Ask separately for the current technical data revision rather than accepting a figure quoted from memory or from a distributor page.

Where the answer is “uninhibited glycol,” record that as the answer rather than as a gap. There are systems and processes for which an uninhibited glycol is exactly what the specification calls for. What must not happen is a request written as though the two were the same line item, priced against each other, and then filled with whichever was cheaper. Alliance does not tell you which your system needs; the equipment owner and the responsible engineer do.

Requirement → evidence → decision boundary

An editorial checklist for your review—not a table of product specifications.

Sort the claim into three layers

A corrosion-inhibitor claim usually arrives as one sentence, and it is doing three jobs at once. The first layer is composition: which additives, at what concentration. The second is test evidence: which bench or loop test, on which metal coupons, for how long, with what acceptance criterion. The third is service evidence: how the fluid behaved in a real system, for how long, under whose monitoring. Ask which layer a given statement belongs to before deciding what it proves.

Composition is frequently proprietary, and a supplier declining to disclose it is not by itself a defect in the evidence. What matters is whether the supplier will state, for the exact product being quoted, which metals the package is formulated to protect, what dilution water it requires, and what the manufacturer’s own monitoring and replenishment guidance says. Record any of those that come back as “not available.”

Say plainly which documents you could not read

The recognised bench methods for coolant corrosion testing are consensus standards published by ASTM, and their full text is sold rather than published. We did not read them for this guide, and we do not paraphrase requirements from documents we have not read. If a supplier cites a test method, the useful follow-up is to ask for the report — the coupon metals, the duration, the temperature, the fluid concentration and the pass criterion — rather than to accept the designation as though it were a result.

The same discipline applies to your own file. NIST describes metrological traceability as a property of a measurement result: a documented unbroken chain of calibrations, a clearly defined property, a full description of the measurement system, a stated value with a documented uncertainty evaluation, and a complete specification of the reference at the time of comparison. Where one of those elements is absent from a corrosion or fluid-condition report, note that it is absent. An undocumented result can still be informative; it should not be filed as though it were traceable.

Bind the claim to the system it has to survive

UFC 3-410-01 couples inhibitor degradation to glycol concentration and instructs designers to coordinate that with the manufacturer. It also requires, in its addition to IMC §1207.2, that systems using an antifreeze solution such as glycol be given “a feed system appropriate for the transfer fluid.” Both are reminders that a protection claim is conditional on how the fluid is diluted, fed and maintained, and that those conditions belong in the same record as the claim.

Wetted materials belong there too, including the ones nobody lists: seals, hoses, pump internals, the makeup-water connection and the heat exchanger. The Hazardous Substances Data Bank entry for propylene glycol notes that the material can be stored in stainless steel, aluminium and lined steel, and that carbon steel is also acceptable “although slight iron contamination” may result. That is a statement about the neat chemical in a storage vessel, not about a diluted, inhibited, circulating system at temperature; treat it as a reason to ask the question, never as the answer.

• Which metals is the package formulated to protect, and which of them are actually in this loop?
• What dilution-water quality does the manufacturer require, and has anyone measured the site water against it?
• What monitoring does the manufacturer specify, at what interval, against which limits?
• Which elements of the supporting report are missing, and who is chasing them?

Requirement → evidence → decision boundary

An editorial checklist for your review—not a table of product specifications.

Both common glycols are colourless

Propylene glycol is recorded as a colourless viscous liquid; the ILO-WHO International Chemical Safety Card aggregated by PubChem describes it as a “COLOURLESS ODOURLESS HYGROSCOPIC VISCOUS LIQUID.” Ethylene glycol is recorded the same way — “Clear, colorless, syrupy, liquid,” with a note that it is a solid below 9 °F. Neither substance is inherently pink, blue, green or yellow.

So any colour in a coolant drum was put there. Dow states that DOWFROST HD is “Dyed fluorescent yellow-green for leak detection purposes.” That is a purpose, attached to one named product, chosen by one formulator. It is not an industry code, and there is no cross-manufacturer convention that lets you read a chemistry, a concentration, an inhibitor type or a service life off a colour. Two fluids that must never be mixed can be the same shade, and one product can ship dyed and undyed.

Identify the fluid from the label, not the sight glass

When a technician reports “the loop is green,” record it as an observation about appearance and leave the identity field unresolved. The identification that matters comes from the fill record, the delivery documents and, where those are missing, an investigation the equipment owner and the fluid supplier define together. Colour is useful for exactly the job Dow assigns it — spotting that something has leaked — and it is useful for noticing that this drum does not look like the last one, which is a reason to ask a question rather than an answer to one.

Where a site standardises on a colour internally, write the convention down, date it, and keep it with the fill records. An undocumented local convention becomes a hazard the moment a contractor who has never seen it makes a top-up decision from it.

Storage is governed by the delivered liquid, not the name

Published closed-cup flash points for the neat glycols are high: propylene glycol is reported at 210 °F (99 °C) closed cup, and ethylene glycol at 232 °F (111 °C) closed cup. OSHA defines a flammable liquid at 29 CFR 1910.106(a)(19) as one with a flash point at or below 199.4 °F (93 °C), sorted into four categories, and adds at (a)(19)(v) that a liquid with a flash point above 199.4 °F which is heated for use to within 30 °F (16.7 °C) of that flash point must be handled as a Category 4 flammable liquid. A hot glycol loop can therefore change the handling question without changing the chemical.

Two details decide whether any of that applies to what you actually received. First, 1910.106(a)(14)(iii) requires that a mixture of compounds with different volatilities be flash-tested “in the form it is shipped” — a diluted or contaminated glycol is not the neat substance. Second, the published figures themselves are method-dependent: the same aggregated record gives propylene glycol 210 °F closed cup, 104 °C by Pensky-Martens closed cup and 225 °F open cup. Classification of a specific delivered liquid is the employer’s determination, made on the exact product’s safety data sheet and the site procedure, not on a value copied from a compound record.

Shelf life is a condition statement, not a date

The aggregated storage guidance for propylene glycol is that it is hygroscopic and “should be stored in a well-closed container, protected from light, in a cool, dry place,” and that at cool temperatures in a well-closed container it is stable. For ethylene glycol the equivalent entry is to store in tightly closed containers in a cool, well-ventilated area away from oxidising agents, at ambient temperature. Hygroscopic matters commercially: an open or badly sealed container drifts in water content, and water content is what the concentration on your purchase order was specifying.

An inhibited fluid adds a second clock, because the inhibitor package is the part with a service life. Ask the manufacturer for storage conditions and any stated shelf life for the exact formulation, put the container condition and the lot identifier into the receiving record, and keep the certificate of analysis with them. Alliance sends a Certificate of Analysis when a customer asks, at no charge — put the request in the purchase order so it arrives with the paperwork rather than months later.

• What does the fill record say the loop contains, independent of what it looks like?
• What storage conditions and shelf life does the manufacturer state for this exact formulation?
• Has the delivered container been open, and is the water content still within the specification?
• Who classifies this liquid for storage, and on which document?

Requirement → evidence → decision boundary

An editorial checklist for your review—not a table of product specifications.