Paul Industries designs and installs animal watering systems and laboratory water systems in Maine. An automated watering system is an unusual piece of engineering: a distribution loop delivering drinking water to thousands of cages, running continuously, with a valve at every cage that is opened by an animal and closed by a spring. It is a water system whose end users cannot report a problem, in a facility where water quality is a research variable as well as a welfare one.
Request a quote or call 201-450-8280
Automated watering: the loop is the whole problem
Maine’s research institutions run colonies large enough that hand-watering with bottles becomes impractical, and automated watering replaces it. The advantages are real: no bottles to fill, autoclave, transport and wash, which removes a substantial labor and cage-wash load. The cost is that the facility now has a permanent water distribution system reaching every rack.
That system has a specific vulnerability. Consumption at any individual cage is tiny, so flow in the distribution is inherently low, and low flow is what lets biofilm establish. Every branch to a rack is a potential dead leg, every valve is a stagnant point, and a rack disconnected for cleaning and not properly capped or removed leaves a stub that will never flush.
The controls are the same principles that govern any high-purity loop, applied to an unglamorous service.
Recirculate rather than dead-end. A loop returning to the treatment plant keeps water moving through the mains, which is far better than a branched dead-end distribution where water sits until an animal drinks. The branches to the racks are short by comparison and that is where the remaining risk concentrates.
Design the branches to be removable. Racks move. A system where a disconnected branch can be physically removed rather than capped avoids the slow accumulation of stubs that degrades these systems over a decade.
Flush deliberately and on a schedule. High-flow flushing of the distribution, including the branches, on a defined interval rather than when someone remembers. Provision for it should be designed in, with flush points at the ends of runs.
Treat the water to suppress growth. Reverse osmosis to remove the dissolved load, then either acidification to a low pH or chlorination at a controlled residual to limit microbial growth in the distribution. Both are established, both are effective, and both are a research variable in their own right, which is the point most worth flagging: acidified water and chlorinated water are not neutral inputs, and a facility changing treatment method should understand that some studies are sensitive to it.
Monitor at the cage, not at the plant. The treatment plant outlet tells you what was produced. A sample at the far end of a distribution loop tells you what an animal actually drinks, and those are different numbers.
Laboratory water: a different specification entirely
The diagnostics and research side of a Maine life-science site needs water of a different kind, specified by grade rather than by potability.
| Water | Typical treatment | What governs it |
|---|---|---|
| Animal drinking water | RO plus acidification or chlorination | Welfare and research consistency |
| Cage wash supply | Softening, sometimes RO for final rinse | Equipment scaling and spotting |
| Laboratory Type III | RO | Glassware rinsing, general use |
| Laboratory Type II | RO with deionization | Buffers, media, general analysis |
| Laboratory Type I | Polished to around 18.2 megohm-centimeter | Sensitive analysis, molecular work |
The practical point for a site running all of these is that they are frequently specified independently by different departments and then built as separate systems that duplicate treatment. A shared pretreatment and RO front end serving several polishing loops is usually cheaper to build and cheaper to run than parallel trains, provided the shared section is sized for the total and the polishing is local to each use.
The counterpoint is that sharing couples the systems. A problem in the common section affects everything downstream, which on a site where an animal facility and a diagnostics operation share a building means a water event becomes a whole-site event. Whether that coupling is acceptable is a genuine decision rather than a technicality.
| Continuous load | Maine per year | At the 8.13 cent US average |
|---|---|---|
| 15 kW | $16,372 | $10,683 |
| 30 kW | $32,745 | $21,365 |
| 60 kW | $65,489 | $42,730 |
At 12.46 cents per kilowatt-hour, 1.53 times the national average (EIA, 2024), the running cost is moderate. The larger consideration on a research site is continuity: an animal facility whose watering system fails has a welfare emergency within a day, which is a stronger argument for redundancy and storage than any energy calculation.
What we build
We design and install pretreatment, reverse osmosis, deionization and polishing, acidification and chlorination dosing with control and monitoring, recirculating distribution to racks with flush provision, laboratory water distribution, storage and the instrumentation and alarms around all of it. Where product-contact or high-purity construction is warranted we build to ASME BPE with orbital welding to AWS D18.1, verified slope and drainability, and passivation to ASTM A967 after cleaning per ASTM A380.
Design and monitoring practice follows the expectations of the Guide for the Care and Use of Laboratory Animals on the animal side, and the applicable ASTM D1193 or clinical laboratory grades on the laboratory side, with the specification agreed against the work actually being done rather than defaulted to the highest grade in the building.
Standards referenced: EIA electricity price data · ASME BPE · ASTM A967 · ASTM A380
Frequently asked questions
Do you build animal watering and laboratory water systems in Maine?
Yes, across Bar Harbor, Portland, Westbrook, Scarborough and statewide: pretreatment, RO, deionization and polishing, acidification and chlorination dosing, recirculating rack distribution with flush provision, laboratory water distribution, storage, and instrumentation and alarms.
What is the main risk in an automated watering system?
Low flow. Consumption at each cage is tiny so distribution flow is inherently low, and low flow lets biofilm establish. Every rack branch is a potential dead leg, every valve a stagnant point, and a disconnected rack that was capped rather than removed leaves a stub that will never flush.
Recirculating or dead-end distribution?
Recirculating. A loop returning to the treatment plant keeps water moving through the mains rather than letting it sit until an animal drinks. The branches to racks remain the residual risk, which is why they should be short and physically removable rather than capped.
Acidification or chlorination?
Both are established and effective at suppressing growth in distribution. The point worth flagging is that neither is a neutral input: acidified and chlorinated water are research variables in their own right, and a facility changing treatment method should know that some studies are sensitive to it.
Where should we sample?
At the far end of the distribution, not at the treatment plant outlet. The plant outlet tells you what was produced; a sample at the end of the loop tells you what an animal actually drinks, and those are frequently different numbers.
Is automated watering better than bottles?
At scale, usually, because it removes the labor of filling, autoclaving, transporting and washing bottles and takes a substantial load off cage wash. The trade is that the facility acquires a permanent water distribution system reaching every rack, with the maintenance and monitoring that implies.
What laboratory water grades will we need?
Typically several: RO for glassware and general use, RO with deionization for buffers and media, and polished water around 18.2 megohm-centimeter for sensitive analysis and molecular work. The grade should be agreed against the work actually being done rather than defaulted to the highest in the building.
Should systems share a front end?
Often yes on cost, with shared pretreatment and RO feeding several local polishing loops sized for the total. The counterpoint is coupling: a problem in the common section affects everything downstream, so on a site where an animal facility and a diagnostics operation share a building, a water event becomes a whole-site event.
How much redundancy does a watering system need?
More than the energy arithmetic would suggest. An animal facility whose watering system fails has a welfare emergency within about a day, so storage and redundancy are justified by continuity rather than by efficiency, and alarms need to reach someone who will act overnight.
How do I get a quote for a Maine water project?
Use the form on this page or call 201-450-8280. Useful inputs are cage or rack count and layout, current watering method, feed water analysis, which laboratory grades are needed and where, and whether existing systems are shared or independent.
What water quality is used for animal drinking water?
Treated water, commonly reverse osmosis product with acidification to a pH around 2.5 to 3 or chlorination to a low residual, to suppress microbial growth in the distribution system while remaining acceptable to the animals. The choice is set by the facility's veterinary and research requirements.
What does acidifying animal drinking water accomplish?
It holds the water at a low pH that suppresses bacterial growth in the distribution system and the cage valves, which is the approach many barrier facilities prefer to chlorination because it does not add a reactive chemical. The pH is set and monitored to keep it within the range the animals tolerate.
What flow and turnover does a recirculating watering loop need?
Continuous circulation at a velocity that prevents stagnation in every branch, with the loop returning to the treatment point so that all water is retreated on each pass, and with rack manifolds designed so that no valve sits on a dead end. The loop is designed like a pharmaceutical water loop at animal room scale.
How are rack manifolds and cage valves kept clean?
By periodic flushing, by sanitising the loop with the manufacturer's recommended chemistry, and by monitoring water at the valve. Cage valves are checked for function because a stuck valve either floods a cage or leaves animals without water.
What does an automated watering system's monitoring have to prove?
That water reached every rack at the required pressure, that the treatment held its setpoint, and that no line went stagnant, with alarms that reach the animal care staff. The monitoring exists to show the animals were watered, which is the one failure the system cannot allow.
How is the watering system sanitised?
By circulating a sanitising solution through the loop and manifolds on a schedule, followed by flushing and verification, with the animals' water supply maintained during the process. Sanitisation is scheduled maintenance and is often neglected.
How is a point-of-use polisher used in a laboratory water system?
As the final stage that takes purified water from the building loop and polishes it to the grade a specific instrument or assay needs, so that the central system does not have to produce the highest grade everywhere. Polishers concentrate the highest cost where it is actually needed.
Should laboratory water be centralised or produced at the bench?
Type II is often centralised with distribution to the labs; Type I is usually polished at the point of use from Type II feed, because Type I degrades in distribution. The split keeps the distributed system manageable.
How does water quality affect research reproducibility?
Water is an ingredient in media, buffers and animal diets, and variation in its composition or microbiology introduces variation into results. Facilities treat water as a controlled variable with records for that reason.
What about water for cage wash and bottle washing?
Cage wash uses softened hot water; bottle filling uses the same treated water as automated watering. The two demands are large and periodic, and the treatment plant is sized for the peaks.
How is water quality maintained through the building?
By materials that do not leach, by circulation, by avoiding dead legs, and by point-of-use filters or polishers where the grade requires. Distribution is where quality is lost, as in any high-purity system.
How does the water system relate to the barrier?
Water crossing into the barrier is treated to exclude organisms, and the treatment is validated as a barrier crossing. A break in water treatment is a break in the barrier.
What happens to a watering system during a power or supply failure?
Reserve storage and a backup pump keep water available to the racks for the period the facility's risk assessment requires, with alarms to the animal care staff so that manual watering can begin before the reserve is exhausted. Animals cannot go without water, so the failure case is designed in.
How is a watering system validated?
By verifying flow and pressure at every rack, treatment performance, microbiological quality at valves, and alarm function, then monitoring through commissioning before animals are housed. The record supports accreditation.
What is the commonest automated watering problem?
Biofilm in the loop from inadequate flushing and sanitisation, detected when valve samples fail long after station samples passed. Sampling at the valve and scheduling sanitisation prevent it.
Sanitizing the loop, which is the maintenance nobody schedules
Continuous treatment suppresses growth in a watering distribution. It does not eliminate what has already established, and over years every one of these systems accumulates biofilm somewhere. Periodic sanitization is what resets that, and it is routinely deferred because the system appears to be working.
Three approaches are used and they suit different installations.
Hyperchlorination. Dosing the loop to a chlorine concentration far above the maintenance residual, holding it for a defined contact time, then flushing thoroughly until the residual returns to normal. Effective, inexpensive, and it demands care at both ends: the concentration must actually reach the far branches rather than just the mains, and the flush must be verified rather than assumed, because residual chlorine left in a drinking water system reaching animals is a welfare and a research problem.
Hot water sanitization. Where the materials and the design permit it, circulating water hot enough to be lethal is thorough and leaves no chemical residue to flush. The constraints are that the distribution, the valves and the seals must all tolerate it, which has to be a design decision rather than a later discovery, and that the system must be drained or cooled before animals have access.
Chemical sanitizers other than chlorine. Used where chlorine is unsuitable, with the same two requirements: full reach and verified removal.
Whichever is chosen, the same practical points apply. Reach is everything, so the procedure should be designed to push sanitant to the ends of every branch, which means flush points at the extremities and a sequence that works through them rather than relying on the mains recirculating. Verification belongs at the far end, sampling where an animal drinks rather than at the plant. And the whole exercise needs a plan for the animals: racks moved to bottles, or a section isolated and done in stages, which is why this gets deferred and why it should be a scheduled operational commitment rather than a good intention.
The design consequence is worth taking seriously at the outset. A system built so that it can be sanitized section by section, with isolation valves and flush points placed for that purpose, is a system that will actually get sanitized. One that requires the entire facility to go to bottles for two days is one that will be sanitized far less often than it should be, and the difference is decided on the drawing rather than in the operating procedure.
Watering or laboratory water questions in Maine?
Tell us your rack count, current watering method and which laboratory grades you need. Sharing a front end is often cheaper, and sometimes the wrong idea. Call 201-450-8280 or use the form below.