Replace or Re-Inhibit Loop Glycol? How to Decide From Fluid Tests

To decide whether to replace or re-inhibit loop glycol, check reserve alkalinity, pH, glycol percent, metals and microbes against the fluid supplier's limits. Re-inhibition suits a clean loop whose glycol percent can be corrected; replacement suits low pH, severe contamination or degraded fluid (OCP, 2026; Dow, 2008).

• The Open Compute Project PG25 guideline lists in-use reserve alkalinity above 4 mL by ASTM D1121 for conventional inhibitor technology, on a 10 mL ready-to-use sample (OCP, 2026).

• Dow states that any DOWFROST fluid with a pH below 7.0 should be replaced (Dow, 2008).

• One seller starts change-out planning when reserve alkalinity falls below 70% of nominal in one article, and schedules change-out at 25% to 50% of baseline in another (Alliance Chemical, 2026).

• ASTM D1121-25 states that reserve alkalinity is not a dependable measure of a coolant's ability to prevent corrosion (ASTM D1121-25, 2025).

This method covers inhibited ethylene or propylene glycol in closed heating, chilled water and liquid cooling loops. It does not cover engine coolants, brines, open systems or uninhibited glycol, and it never overrides the named fluid supplier's own limits.

When should you decide between replacing and re-inhibiting loop glycol?

You should decide between replacing and re-inhibiting loop glycol whenever a test result falls outside the fluid supplier's in-use limits, after a contamination event, and at each scheduled laboratory analysis. The Open Compute Project (OCP) guideline for propylene glycol fluids in liquid cooled racks, Revision 2.0, sets annual testing of each loop by a qualified lab as a minimum, covering glycol concentration, pH, reserve alkalinity, appearance, inhibitor levels, chloride, sulfate, hardness, degradation acids and dissolved metals (OCP, 2026). When an on-site test shows an out of specification result, OCP says a sample should go to a qualified lab within one week (OCP, 2026). Dow's 2008 guide recommends yearly samples once a system is in operation and offered free laboratory analysis for systems holding 950 L (250 US gal) or more of its fluid, with on-site checks for smaller systems (Dow, 2008). A decision point is not the same as a replacement trigger. OCP states that deviation outside its ranges does not necessarily warrant fluid replacement, and it sends users to the fluid supplier for formulation and site specific guidance (OCP, 2026). Its replacement rule is broad: replace the fluid at the end of its useful life as the supplier determines, or when it no longer meets the supplier's performance specifications (OCP, 2026). The decision therefore starts from the supplier's limits for the named product, then uses the criteria in this guide to choose between keeping, re-inhibiting and replacing.

What glycol fluid test data do you need for the decision?

You need the glycol fluid identity, its fill baseline, a recent laboratory panel and the loop's maintenance history before you can choose between re-inhibiting and replacing. Identity means the product name, the glycol type and the inhibitor technology, because OCP gives different pH ranges for conventional inorganic and organic acid inhibitor fluids (OCP, 2026). The baseline is the certificate of analysis (COA) and an as-received sample; OCP advises retaining a sample from every bulk delivery for a minimum of 6 months, and to keep a fresh-fluid sample and a start-up sample for the life of the fluid (OCP, 2026). One seller's guide makes the same point about reserve alkalinity: without a documented baseline, later results are interpretations rather than percentages (Alliance Chemical, 2026). The laboratory panel should report glycol concentration by ASTM E202 or ASTM D3321, pH by ASTM D1287, reserve alkalinity by ASTM D1121, metals, chloride and sulfate by ASTM D6130 and ASTM D5827, and appearance (OCP, 2026). Degradation acids such as formate and glycolate can be measured by ion chromatography, which ASTM D8528-25 covers for engine coolants (ASTM D8528-25, 2025). History closes the picture: OCP's record table asks for fluid analysis reports, component replacements and total system volume, which is needed to calculate inhibitor additions (OCP, 2026). Add the makeup water volume and every top-up or additive since the last sample, since each one changes how the results read. Take the sample from well-mixed, circulating fluid, not from an expansion tank, low-point drain or dead leg, and flush the port first (OCP, 2026).

Which glycol test results mean the fluid must be replaced?

Glycol test results point to replacement when pH falls below the supplier's replacement level, when contamination is severe, or when the fluid shows clear degradation; a single low reserve alkalinity result usually points to re-inhibition first. OCP values are written for PG25 in liquid-cooled racks; for other loops, read them beside the fluid supplier's own limits. The criteria table below decides, test by test, whether a result keeps the fluid in service, points to re-inhibition or points to replacement.

The pH row carries the clearest replacement rule. Dow states that any DOWFROST fluid with a pH below 7.0 should be replaced, and that a pH between 7.0 and 8.0 can be adjusted (Dow, 2008). Reserve alkalinity should not condemn fluid on its own, because ASTM D1121-25 warns that reserve alkalinity is not a dependable measure of a coolant's ability to prevent corrosion, nor of its remaining life (ASTM D1121-25, 2025). Chem-Aqua, a water treatment supplier, lists pH depression, a sharp aldehyde or septic odor, severe steel corrosion with high copper levels, and microbiological problems as symptoms of glycol degradation, and notes that salvaging degraded glycol may not be practical, depending on severity (Chem-Aqua, 2018). When two or more of those symptoms appear together, replacement is the stronger case (inference). Copper and iron need a trend, not one point: for results above 5 ppm, OCP's response is to monitor more frequently to determine whether levels are rising (OCP, 2026). A rising metal result alongside falling pH suggests corrosion the inhibitor has not held (inference).

When can an inhibitor booster restore a glycol loop?

An inhibitor booster can restore a glycol loop when the glycol is the right product at a correctable concentration, pH sits above the replacement level, metals are steady, and the additive is approved by the original fluid supplier. OCP is explicit on the last point: when analysis shows depletion, use only additives approved by the original fluid supplier, because unapproved additives or overdosing can precipitate inhibitor compounds (OCP, 2026). Conventional closed system treatment does not fit glycol; Chem-Aqua states that conventional nitrite or molybdate closed system programs will not adequately protect glycol systems, and gives inhibitor recommendations based on a laboratory analysis (Chem-Aqua, 2018). One seller warns that adding a silicate-containing formulation to a loop whose pH has dropped below 7.5 can precipitate silica gel, so test before dosing (Alliance Chemical, 2026). Where pH needs raising, Dow allows adjustment between pH 7.0 and 8.0 with a 50% solution of sodium hydroxide or potassium hydroxide (Dow, 2008). Those are corrosive solutions: before handling them, read Section 8 of the product's Safety Data Sheet (SDS) for protective equipment, which 29 CFR 1910.1200(g) requires the supplier to provide in the US (OSHA, 2012) or REACH Article 31 in the EU (Regulation (EC) No 1907/2006, Article 31), and dose only under the fluid supplier's guidance (OCP, 2026). After dosing, circulate before testing again; Dow's fill procedure calls for at least 24 hours of circulation to ensure complete mixing (Dow, 2008). A booster cannot reverse damage that metals results already show (inference).

How do contamination and microbes change the glycol replacement decision?

Contamination and microbes shift the glycol replacement decision toward replacement, because a booster restores inhibitor but cannot remove foreign fluid, debris, biomass or the acids microbes produce. OCP sends any sample with particulates or cloudiness to a qualified lab, allows sidestream filtration depending on the contaminant, and states that fluid replacement may be necessary after severe contamination (OCP, 2026). Dilution is one route to microbes. OCP warns that bio-growth may occur if propylene glycol concentration is not kept at about 25% or above (OCP, 2026), and Dow states that glycol solutions below 25% may be at risk for bacterial contamination (Dow, 2008). Chem-Aqua describes the worst case, a small amount of glycol entering a water loop, for example from incompletely flushed coils: bacteria use the glycol as a food source and produce acidic waste that drives loop pH well below 8, after which old iron corrosion products dissolve, the water turns black and a septic odor develops (Chem-Aqua, 2010). Water quality contaminants matter too: Dow notes that more than 25 ppm of chlorides, sulfates or ammonia could contribute to system corrosion (Dow, 2008). For microbial counts, BSRIA BG 50/2021 defines total viable count (TVC) as a measure of aerobic bacteria expressed as cfu/ml (Designing Buildings, 2021), but no in-use glycol limit for TVC appeared in the sources checked. A positive count with normal pH and appearance points to correcting concentration and retesting; a positive count with low pH, odor and dark fluid points to cleaning and replacement (inference).

Why do published glycol replacement thresholds differ?

Published glycol replacement thresholds differ because sources measure reserve alkalinity in different ways, write for different inhibitor chemistries, and separate new fluid from in-use fluid differently. Reserve alkalinity is the clearest case. OCP states an absolute value, above 4 mL by ASTM D1121 on a 10 mL ready-to-use sample for conventional fluids, and "to report" for organic acid fluids (OCP, 2026). Chem-Aqua lists a reserve alkalinity range of 3.0 to 6.0 for glycol loops without stating the unit or method (Chem-Aqua, 2018). One seller instead uses percent of baseline and is not consistent across its own articles: one marks 70% of nominal as a change-out planning trigger and calls for flush and recharge below 30% with pH below 7.4 and active corrosion, another schedules change-out at 25% to 50% of baseline (Alliance Chemical, 2026). Inhibitor chemistry explains part of this. ASTM D1121-25 notes that some inhibitors contribute little reserve alkalinity yet protect metals well, so the magnitude of reserve alkalinity is not always a good criterion of protection (ASTM D1121-25, 2025). New and used fluid also differ: OCP states that its new and in-use PG25 acceptance criteria differ by design (OCP, 2026), and ASTM D8039-16(2023) specifies concentrated and prediluted HVAC heat transfer fluids as supplied, so its requirements describe fresh fluid rather than in-service limits (ASTM D8039-16(2023)). Units also hide real differences. OCP lists copper below 3 ppm in use (OCP, 2026), while the seller treats copper above 300 ppb, or 0.3 ppm, as active corrosion on a degrading loop (Alliance Chemical, 2026), so once converted, the seller's level is one tenth of OCP's limit.

Glycol replacement decision worked example (illustrative)

A worked example shows how the criteria table turns a glycol lab report into a re-inhibit decision. The loop and every result in the table below are illustrative, set against the OCP in-use limits for a conventional PG25 fluid.

The illustrative loop is a 12,000 L chilled water loop filled four years earlier, with a fill baseline of 7.0 mL reserve alkalinity. Its current 3.2 mL is 46% of that baseline (3.2 divided by 7.0, illustrative). The loop was filled at 26% glycol (illustrative). Dilution from 26% to 22% would lower reserve alkalinity only to about 5.9 mL (7.0 x 22/26, illustrative), so most of the loss to 3.2 mL is inhibitor consumption. With pH, metals and appearance inside limits, the fluid shows depletion plus dilution but no sign of degradation (inference). The decision is to re-inhibit: raise concentration with a supplier-approved concentrate, add supplier-approved inhibitor under the supplier's guidance (OCP, 2026), circulate, and retest. Find and stop the makeup source first, or the correction will not hold. Resample after circulation and confirm that glycol concentration is back in the 24% to 29% band and reserve alkalinity above 4 mL before closing the work order (OCP, 2026). The same report reads differently against seller rules: 46% of baseline falls inside one seller's 25% to 50% change-out band and below its other article's 70% planning trigger, yet above the 30% level where that article calls for flush and recharge (Alliance Chemical, 2026). That gap is why the decision cites supplier limits. Had the illustrative pH been 6.8 with a burnt odor, Dow's guidance would call for replacement instead (Dow, 2008).

How do you present a glycol replacement decision to a building owner?

Present a glycol replacement decision to a building owner as one page that states the decision first, then the results against named limits, the trend, the options with their risks, and the date of the next check. Open with one sentence the owner can approve, such as "Re-inhibit loop 2 now and retest in four weeks", then show each result beside its limit and source, as in the worked example. Show the trend, not only the latest number, because a slow decline and a sudden drop call for different responses. List the alternatives honestly: re-inhibiting, replacing, and deferring, with what each risks and needs, such as downtime, disposal of spent fluid under SDS Section 13 and local waste rules (OSHA, 2012), and a flush. State what the decision cannot promise. OCP notes that fluid suppliers typically do not guarantee the performance or lifespan of the fluid (OCP, 2026), so the page should say how success will be checked. Attach the laboratory reports, the COA and the records OCP lists, including fluid analysis reports, maintenance history and system volume (OCP, 2026). Name who acts and by when for each step. Some equipment makers may require routine fluid analysis as part of warranty conditions (OCP, 2026), so record whether a warranty is involved. Lubinpla can draft this kind of decision summary from your own lab reports, TDS, SDS and site history, with each result cited, for your review before it goes to the owner.

Where the glycol replacement method does not apply

The glycol replacement method in this guide does not apply to engine coolants, organic acid fluids judged by reserve alkalinity alone, unidentified or mixed fluids, and loops under a manufacturer's own fluid program. Engine cooling systems use a different family of specifications; ASTM D8039-16(2023) states that its HVAC heat transfer fluids are not appropriate for systems with internal combustion engines (ASTM D8039-16(2023)). For organic acid inhibitor fluids, OCP lists reserve alkalinity as "to report" (OCP, 2026), so the reserve alkalinity rows of the criteria table do not decide those loops; inhibitor concentration from the supplier's own test does (inference). An unidentified glycol or a loop topped up with several products needs identification first, because each supplier's limits assume its own formulation and OCP advises using the same heat transfer fluid throughout to avoid introducing different fluids (OCP, 2026). Loops below about 25% glycol, kept dilute for heat transfer reasons, sit outside the inhibitor assumptions of products such as Dow's; Dow warns that diluting below 25% to 30% as glycol may reduce the inhibitor to a level that may not protect against corrosion (Dow, 2008). Food, potable water or pharmaceutical duties add product and regulatory requirements that this guide does not cover. Where a warranty requires the equipment maker's fluid analysis (OCP, 2026), that program's limits govern.

Key takeaways

• Start from the named fluid supplier's in-use limits; the criteria table orders the checks.

• pH below the supplier's replacement level, severe contamination or clear degradation points to replacement.

• Low reserve alkalinity with clean, steady fluid points to re-inhibition with a supplier-approved additive.

• Low glycol percent usually means dilution: fix the makeup source before correcting it.

• Published thresholds differ by method, inhibitor chemistry and units, so cite the source beside every limit.

Lubinpla applies this kind of reasoning to your own documents and site history and drafts the document for your review before it goes out.

Lubinpla does not sell chemical products and names no brands in recommendations. Our reviewer also leads a specialty chemicals distribution company.

Sources checked 2026-10-14.

Sources

1. Open Compute Project. Guidelines for Using Propylene Glycol-Based Heat Transfer Fluids in Liquid Cooled Racks, Revision 2.0, Version 1.0.0. September 2026. https://www.opencompute.org/wp-content/uploads/2026/10/OCP-PG25-Guidelines-V1.0.0.pdf

2. ASTM International. ASTM D1121-25 Standard Test Methods for Reserve Alkalinity of Engine Coolants and Antirusts. 2025. https://store.astm.org/d1121-25.html

3. ASTM International. ASTM D8039-16(2023) Standard Specification for Heat Transfer Fluids (HTF) for Heating and Air Conditioning (HVAC) Systems. 2023. https://store.astm.org/d8039-16r23.html

4. ASTM International. ASTM D8528-25 Standard Test Method for Determination of Fluoride and Glycol Degradation Products in Engine Coolants by Ion Chromatography. 2025. https://store.astm.org/d8528-25.html

5. Dow. Engineering and Operating Guide for DOWFROST and DOWFROST HD Inhibited Propylene Glycol-based Heat Transfer Fluids, Form No. 180-01286-0208. 2008. https://cdn.shoplightspeed.com/shops/621040/files/64538514/dowfrost-inhibited-propylene-glycol-heat-transfer.pdf

6. Chem-Aqua. Technical Bulletin 3-004, Protecting Glycol-Water Closed Systems. 2018. https://www.chemaqua.com/fr-ca/wp-content/uploads/sites/5/2024/07/TB3-004.pdf

7. Browning, A. and Berry, D. (Chem-Aqua). Selecting and Maintaining Glycol Based Heat Transfer Fluids. FMLink, 2010. https://www.fmlink.com/selecting-and-maintaining-glycol-based-heat-transfer-fluids/

8. Alliance Chemical. Inhibitor Depletion Lifecycle and Coolant-as-a-Service. 27 April 2026. https://alliancechemical.com/blogs/articles/inhibitor-depletion-lifecycle-coolant-as-a-service

9. Alliance Chemical. Reserve Alkalinity Testing for Data Center Coolant Loops. 27 April 2026. https://alliancechemical.com/blogs/articles/reserve-alkalinity-testing-cadence-astm-d1121

10. Designing Buildings. Total viable count (definition from BSRIA BG 50/2021). 2021. https://www.designingbuildings.co.uk/wiki/Total_viable_count

11. US Occupational Safety and Health Administration. Hazard Communication Standard: Safety Data Sheets, OSHA 3514. 2012. https://www.osha.gov/sites/default/files/publications/OSHA3514.pdf

12. European Union. Regulation (EC) No 1907/2006 (REACH), Article 31 Requirements for safety data sheets, consolidated text 02006R1907-20250623 (EUR-Lex). 2025. https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX:02006R1907-20250623

Written by the Lubinpla team from the sources listed and checked against them before publication.

Reviewed by Jonghwan Moon, Founder & CEO, Lubinpla, 2026-10-14.