Closed Loop Water Treatment Log Sheet With Ranges and Actions

This closed loop water test log suits sealed heating, chilled water and glycol loops sampled by hand at set visits. It does not fit open cooling towers, steam boilers or sprinkler systems, and it does not replace the treatment specialist's interpretation or a laboratory report.

What is a closed loop water treatment log sheet for, and when do you use it?

A closed loop water treatment log sheet is a one-row-per-result record that lets a technician, facilities manager or water treatment specialist act on each closed loop test the day it is taken. Use it from the first fill through every service visit on a sealed heating, chilled water or glycol loop. BS 8552:2012 gives recommendations on "the location and suitability of sampling points and frequency of sampling and recordkeeping" for closed systems (BSI, 2012), but the published scope also states: "This standard does not provide guidance on the interpretation of results or associated action levels." (BS 8552:2012, scope). The log closes that gap at the site level. Each row carries the reference range and the name of its source before the result is written, the sample point and method that produced the number, and the action taken when the number falls outside the range. A result kept that way can be compared with the next visit, handed to a new contractor, or sent to the fluid supplier without a phone call to explain it. The hydronic glycol and pH log templates at the top of search results record a glycol reading, a pH reading and a dose, and leave every target to the vendor. The editable Excel log and a CSV of the same header row are at the end of this page.

Which columns does the closed loop water test log have?

The closed loop water test log has fifteen columns, and every result row fills all of them so the row reads correctly on its own. The table below shows each log column, what goes in it, its unit and an example value; all example values are illustrative and come from no real site.

In the illustrative pH row, the result of 7.6 is written beside the 8.0 to 10.0 range and its source, so the N in the next column needs no explanation.

The closed loop water test log uses one row per parameter rather than one row per visit, for three reasons. First, each parameter has its own method, unit and range source, and a wide visit row cannot hold all of them without losing which number came from which kit. Second, a filter on Loop ID and Parameter gives a trend for any parameter across visits without rearranging the sheet. Third, a laboratory result that arrives two weeks after the visit is added as its own row with the laboratory method in the Method column, not typed over the on-site reading. The T value in the In range column marks a result that is inside the published range but has moved away from the loop's own fill baseline, which matters most for conductivity and reserve alkalinity. The Excel file has an Instructions sheet, the blank Log sheet with a frozen header row and an Example sheet marked illustrative; the CSV holds the Log header row only.

Where do the reference ranges in a closed loop water test log come from?

The reference ranges in a closed loop water test log come from the fluid or inhibitor supplier, the site water treatment specification and, for glycol loops, published operating guides, never from the result itself. BS 8552:2012 names no action levels, and BSRIA BG 50/2021, the second edition of the BSRIA guide published in November 2021 (Designing Buildings, 2021), "is a guide and not a standard or regulation" (PBC Today, 2022). The table below lists published ranges the log can carry, each with the condition under which its source gives it.

Open Compute Project marks its in-use values as examples and refers the reader to the fluid manufacturer, so the copper row reads below 3 ppm only for in-use PG25 in liquid-cooled systems (Open Compute Project, 2026). Water-only loops take their ranges from the treatment specialist's programme. Where two sources give different ranges for the same parameter, as Dow and Open Compute Project do for pH, the log uses the range of the fluid actually in the loop and names that source; it does not average them or pick the wider one. A loop filled with an organic acid inhibitor fluid and logged against a conventional fluid's pH range will show false alarms at every visit.

How do you fill in a closed loop water test log at each visit?

A closed loop water test log is filled in seven steps at each visit, and the first step happens before any result exists.

1. Write each parameter's reference range and range source into the log before the visit, and keep them unchanged for that round; a range changed after a result is seen starts a new round.

2. Record the fill baseline once: Open Compute Project recommends taking a sample within the first 1 to 3 days of operation and keeping it for the life of the fluid (Open Compute Project, 2026).

3. Draw the sample from well-mixed, circulating fluid, not from expansion tanks, reservoirs, low-point drains or dead legs, and flush the port first (Open Compute Project, 2026). Wear the protection named in section 8 of the fluid's safety data sheet, which the US Hazard Communication Standard describes as exposure controls and personal protection (OSHA, 2012); hot or pressurized loops follow the site's permit procedure.

4. Label the bottle with the date, loop and location, using the same names at every visit so results can be trended (Open Compute Project, 2026).

5. Test, then write the result with its unit and the method or kit used.

6. Mark Y, N or T against the range written in step 1, write the action, and send a laboratory sample within one week of an out-of-specification on-site result (Open Compute Project, 2026).

7. Read the makeup water meter and enter the volume added since the last visit, in litres.

How do you read a closed loop water test result that is out of range?

A closed loop water test result that is out of range is read against the range source written in the same row, then against the previous visits, and only then acted on. The table below gives the meaning and the first action for each parameter; the thresholds are those of the cited source and its condition.

The row most logs miss is the last one: closed loops typically need less than 5 percent makeup a year, and NIH attributes larger losses to minor leaks in mechanical seals or overflow from expansion tanks (NIH Office of Research Facilities, 2023). Dow names a pH below 7.0 as the point where inhibited propylene glycol fluid should be replaced rather than adjusted (Dow, 2008). Before any action, check the row itself: a result from a different sample point, an uncalibrated meter or a new kit lot is retested first, because acting on a measurement error costs more than a second reading. Any adjustment chemical goes in only on the fluid supplier's instruction and under its safety data sheet.

How well does a closed loop water test log predict corrosion in the loop?

A closed loop water test log predicts corrosion only indirectly, because the parameters it records are signs of the conditions that cause corrosion, not a measured corrosion rate. Dissolved oxygen is the clearest gap: BSRIA BG 50/2021 is reported to state that "The level of dissolved oxygen is the most important factor related to corrosion," and that laboratory sampling is not suitable for testing it (H&V News, 2021). A grab sample drawn into a bottle and tested later cannot record dissolved oxygen faithfully, so a log built from grab samples should not claim to.

Reserve alkalinity has a similar limit. ASTM D1121-25 states: "the reserve alkalinity of an engine coolant solution is not a dependable measure of its ability to prevent corrosion, nor can it satisfactorily indicate the additional life of the solution" (ASTM D1121-25, section 5.7). The log therefore reads reserve alkalinity as a trend against the fill baseline, and one vendor guide notes that without a documented baseline, later measurements are interpretations rather than percentages (Alliance Chemical, 2026).

Rising copper or iron is the nearest the log comes to a corrosion signal, and it still lags the metal loss. Corrosion coupons, which NIH describes as metal strips left in the system for 90 days and then analysed for corrosion rate (NIH Office of Research Facilities, 2023), or inline monitoring measure corrosion more directly and belong beside the log.

Closed loop water test log mistakes that hide a problem

Closed loop water test log mistakes that hide a problem usually come from rows that cannot be compared with each other. The common ones are below.

• Writing the range after the result. A range chosen once the number is known will usually fit the number; fix it before the visit.

• Leaving out the sample point. A reading from an expansion vessel and a reading from the return header are different measurements, and Open Compute Project advises against expansion tanks and dead legs as sample points (Open Compute Project, 2026).

• Changing the method between visits. A refractometer reading compared with a laboratory glycol result, or pH paper compared with a calibrated meter, shows a step change that may be the method, not the loop.

• Dropping the unit. Conductivity in µS/cm and in mS/cm differ by a factor of 1,000, and hardness as CaCO3 is not the same number as hardness as calcium.

• Ignoring makeup. Losing 25 percent of system volume a month is not uncommon in closed loops, by the NIH account, and makeup water brings in minerals, gases and microorganisms (NIH Office of Research Facilities, 2023). A falling glycol percentage with a rising makeup volume is one finding, not two.

• Treating "in range" as "healthy". Chem-Aqua lists pH depression, aldehyde or septic odor and high copper as signs of glycol degradation (Chem-Aqua, 2018); a row can sit in range while the trend across rows says otherwise.

When does a closed loop water test log not fit the system?

A closed loop water test log does not fit systems that BS 8552:2012 excludes or that need continuous measurement rather than visit-based sampling. The BS 8552:2012 scope listing excludes open circuit systems such as cold water services, open cooling towers and cooling ponds, steam boiler plant and sprinkler systems (BS 8552:2012, scope), and each of those has its own standards and control limits. Glycol loops with an equipment maker's own fluid limits, such as liquid-cooled rack systems, should carry that maker's or the fluid supplier's ranges in the Reference range column instead of the published examples on this page. Loops where dissolved oxygen is the main concern need inline monitoring, since BSRIA BG 50/2021 is reported to treat laboratory sampling as unsuitable for dissolved oxygen (H&V News, 2021). The log is also not a laboratory report: microbiological counts, metals by spectroscopy and degradation products come from the laboratory's own report, which the Action taken column references by report number. Lubinpla's work log bulk import accepts txt, md, csv, pdf and image files but not Excel, so a filled log goes in saved as CSV with the same header row or as PDF. Where a result needs a decision on treatment, replacement or cleaning, the log is evidence for the water treatment specialist, not the decision itself.

Key takeaways

• Write the reference range and its source in the row before the result.

• Bind each result to its sample point, method and unit.

• Read reserve alkalinity and conductivity against the fill baseline.

• Record makeup volume at every visit; closed loops typically need less than 5 percent a year (NIH Office of Research Facilities, 2023).

• Use the log as evidence for the treatment specialist, not as the verdict.