Paul Industries designs, installs and validates high-purity water systems across Indiana: USP Purified Water, deionized water specified by resistivity, and Water for Injection where a product genuinely requires it. Indiana is the state where the WFI default costs the most unnecessary money. Its dominant sector is medical device, and device cleaning validation targets machining residues rather than a parenteral endotoxin limit, so the expensive hot loop most pharmaceutical contractors reach for is usually the wrong answer here.

Request a quote or call 201-450-8280

Dominant requirement DI by resistivity or USP Purified Water, not WFI
Why Device cleaning validation targets machining residues, not endotoxin
Where WFI does apply Combination products carrying a drug, and bloodstream-contact cases
Final rinse The stage that decides whether cleaning validation passes
Industrial power 8.15 cents/kWh, essentially the US average of 8.13 (EIA, 2024)
Reviewed September 2026, against current state code and EIA 2024 energy data

Resistivity, not endotoxin, is the Indiana specification

Pharmaceutical water and device water are specified against different things, and in a device state that difference is worth several hundred thousand dollars over the life of a system.

Compendial pharmaceutical water is specified by monograph: USP 643 for total organic carbon, USP 645 for conductivity as a three-stage test, and for Water for Injection an additional endotoxin limit of 0.25 EU per milliliter. That endotoxin limit is what drives the hot circulating loop, the sanitization regime and most of the operating cost of a WFI system.

Device rinse water is usually specified differently. Deionized water is specified by resistivity, commonly with limits on total organic carbon and particulates, because what the cleaning validation has to demonstrate is removal of cutting fluid, polishing compound, machining debris and handling contamination. Endotoxin is generally not the governing limit, because the product is not going to be injected and will in most cases be terminally sterilized.

The error that follows is predictable. A contractor whose experience is pharmaceutical proposes WFI and a hot loop, the client assumes that must be the safe choice, and an Indiana device plant ends up with a system carrying capital cost it did not need and a permanent running cost it will pay every year. On the numbers below, a hot loop against an ambient system is roughly $7,500 a year in this state, indefinitely, for a specification the validation never asked for.

Where WFI genuinely does apply is worth stating plainly: combination products carrying a drug, devices with bloodstream contact where the risk assessment demands it, and any product where the manufacturer’s own risk analysis concludes the endotoxin limit is appropriate. Those cases exist and should be built properly. They are the minority of Warsaw output.

Three water grades, three different jobs

Water grades found on an Indiana device and pharmaceutical site
GradeSpecified byTypical useLoop
Deionized waterResistivity, plus TOC and particulate limitsUltrasonic cleaning, final rinse of componentsAmbient, recirculated, often with UV
USP Purified WaterUSP 643 TOC, USP 645 conductivityWhere a compendial grade is required but endotoxin is notAmbient with ozone, or hot
Water for InjectionUSP 643, USP 645, plus 0.25 EU/mL endotoxinCombination products, drug-carrying devicesHot circulating, usually
Annual electricity cost of a continuously running loop in Indiana at 8.15 cents/kWh
Continuous loadPer yearOver ten years
4 kW ambient loop$2,856$28,560
15 kW hot loop$10,709$107,090
Difference, 10.5 kW$7,496$74,960

The final rinse is where cleaning validation is won or lost

On a device line the last thing to touch the part before drying and packaging is the final rinse, and its quality determines whether residue testing passes. That makes the rinse loop part of the cleaning process rather than a utility connection made later by someone else.

The design points that matter are how the rinse water is delivered to the part, whether the delivery point can itself be cleaned, whether the water drains completely rather than pooling on the component, and where the sampling point sits so that what is tested is what the part actually saw. A specification that names a resistivity and stops has not addressed any of those.

Frequently asked questions

Do you install water systems in Indiana?

Yes, including the Warsaw orthopedic cluster, Indianapolis and statewide. We deliver deionized and USP Purified Water systems, and WFI where a product genuinely requires it, with generation, storage, distribution, passivation and qualification held in a single scope. On device work that matters because the specification argument on this page, about what the cleaning validation genuinely requires, has to be settled by whoever is accountable for the finished system rather than by a skid vendor quoting against an assumption.

Does a medical device line need WFI?

Usually not. Device cleaning validation targets removal of cutting fluid, polishing compound, machining debris and handling contamination, not control of a parenteral endotoxin limit. Deionized water specified by resistivity, or USP Purified Water, is generally sufficient. WFI becomes relevant for combination products carrying a drug, or where a risk assessment concludes the endotoxin limit applies.

What does specifying WFI unnecessarily actually cost?

The capital difference plus a permanent operating cost. At Indiana’s 8.15 cents per kilowatt-hour, a hot loop against an ambient system on an assumed 10.5 kW continuous difference is about $7,496 a year, roughly $74,960 over ten years, before counting the additional sanitization downtime and the derouging a hot loop eventually needs. That is money spent on a limit the validation never required.

How is DI water specified?

Primarily by resistivity, commonly with limits on total organic carbon and particulates appropriate to the cleaning it supports. Unlike a compendial monograph, the limits are set by what the process requires rather than by a pharmacopoeia, which means the specification has to be justified against the cleaning validation rather than adopted from a template.

Why does the final rinse matter more than the wash?

Because it is the last thing to contact the part before drying and packaging, so its quality determines what residue testing finds. If final rinse water is delivered through a fitting that cannot itself be cleaned, or pools on the component rather than draining, the rinse can add contamination rather than removing it. The rinse loop is part of the cleaning process, not a utility connection.

Where should the sampling point be?

Where it represents what the part actually saw, which is at or as close as practical to the point of delivery rather than at a convenient position on the loop. A sample taken upstream of the final delivery fitting reports the quality of water that had not yet passed through the last thing it touches, which is precisely the component most likely to be contributing a problem.

Do you build ultrasonic and parts-cleaning lines?

Yes, including the water, drainage, ventilation and controls around them. The part most often underestimated is the rinse side: final rinse water quality, how it is delivered, how it drains and how it is sampled will determine whether the cleaning validation passes. We specify the rinse loop as part of the line rather than leaving it as a connection made later.

Can you upgrade an existing DI system?

Frequently. Common upgrades are improving the polishing stage to hold resistivity more consistently, adding UV for microbial and TOC control, correcting loop dead legs that cause quality to drift at particular points of use, and improving sanitization access. The diagnosis worth doing first is sampling at the generation outlet and at points of use separately, which separates a generation problem from a distribution one.

Does Indiana energy cost affect the design?

Very little, which is useful. At 8.15 cents per kilowatt-hour against a national average of 8.13 (EIA, 2024), Indiana is essentially at the midpoint. The water grade decision here should follow the cleaning validation rather than the tariff. What the tariff does mean is that an unnecessary hot loop costs a predictable, unremarkable amount every year rather than a dramatic one, which is exactly why it goes unnoticed.

How do I get a quote for an Indiana water system?

Use the form on this page or call 201-450-8280. The most useful first answer is what your cleaning validation actually requires of the water, because that determines the grade and everything downstream. After that we need your peak draw rather than a daily average, how many delivery points the system feeds, a current feed water analysis, and whether ultrasonic cleaning lines form part of the scope.

What water does a device rinse line actually need?

Water pure enough that it leaves nothing behind on a surface whose cleanliness is a product attribute, which for final rinse usually means deionised water at a specified resistivity with organics controlled. The specification should be derived from what residue matters to the product, not adopted from a pharmacopoeia written for a different purpose.

Does a medical device line need Water for Injection?

Very rarely, and specifying it is the most expensive default in this state. Water for Injection exists to control endotoxin for parenteral products. A terminally sterilised device with an endotoxin limit established by testing is generally well served by high-quality deionised or purified water, verified by the device’s own endotoxin results rather than by the water’s compendial grade.

What does specifying it unnecessarily cost?

A generation plant that is more complex than required, a distribution system that must be held hot or otherwise controlled continuously, a qualification and monitoring programme with sampling and testing costs that continue for the life of the facility, and a maintenance obligation. The capital premium is the smaller part; the recurring cost is what compounds.

How does water contribute to device endotoxin results?

Directly, because endotoxin from organisms in the rinse water deposits on the part and survives drying. A device failing its endotoxin limit with clean bioburden results is frequently being contaminated by its own final rinse. That is why microbial control in a device water system matters even when the water is not compendial and no microbial specification is written.

How should rinse tanks be arranged?

As a cascade, with parts moving from the dirtiest rinse to the cleanest and water flowing the opposite way, so fresh water contacts the cleanest parts and the carried-over soil is progressively diluted. Counter-current arrangement dramatically reduces water consumption for the same final cleanliness compared with a series of independently fed tanks.

How is water consumption reduced on a rinse line?

Counter-current cascading, controlling drag-out by allowing parts to drain between tanks, using flow only when parts are present rather than continuously, and monitoring rinse tank quality so fresh water is added on condition. Rinse lines running continuous overflow around the clock are common, and the consumption is usually many times what the process needs.

What quality does the ultrasonic tank need?

Better than people assume, because the ultrasonic bath is where the soil accumulates and where redeposition happens. A bath running past its useful life redeposits onto parts the soil it removed from earlier ones. Bath monitoring, defined change frequency and adequate filtration matter as much as the transducer power that gets specified so carefully.

Can microbial growth occur in a deionised system?

Readily, particularly at ambient temperature with low flow, and it matters here because the organisms contribute both particulate and endotoxin to a final rinse. Ultraviolet treatment, maintained flow, and control of stagnant branches are the usual measures. A system monitored only on resistivity can be microbiologically poor while reporting excellent numbers.

How is a device water specification set without a pharmacopoeia?

By identifying what residue would fail the product, then setting limits on the attributes that control it: resistivity for ions, organic carbon where it matters, microbial and endotoxin where the device’s own testing depends on it, and particulate for cleanliness. Each limit should have a reason, which is what makes the specification defensible in an audit.

Should the water system be validated?

If water quality affects product quality, the system should be qualified and controlled proportionate to that risk, whatever the framework. That usually means installation and operational qualification, a performance period establishing behaviour, defined monitoring and change control. The depth is lighter than a compendial system and the principle is the same.

Does a process water system need change control after handover?

Yes, and its absence is the most common finding we see on device water systems. Filters get changed to different types, branches get added, lamps get replaced with different ratings, and eventually the installed system is not the one that was qualified. When a cleanliness problem appears, nobody can say what changed or when.

How should sampling points be arranged?

At the points of use that actually feed critical rinses, plus the loop return, with ports designed so the sample describes the water rather than the port. On a device line the most informative sample is usually taken at the final rinse tank feed, because that is the water that determines what remains on the product.

Can an existing deionised system be upgraded?

Usually, and in stages. Adding point-of-use polishing, ultraviolet treatment, better monitoring and removing stagnant branches addresses most of the problems that lead to inconsistent rinse results, without replacing the generation plant. Establishing where the current system actually fails, by measuring, is worth doing before specifying anything new.

Does Indiana feed water affect the design?

Where plants draw on groundwater, which is common across much of the state, hardness and iron content are higher than surface supplies and the pretreatment has to address them before the deionisation stage. A design based on a softened municipal supply will behave differently on a harder source, usually appearing as shortened resin life or scaling.

How is monitoring instrumentation kept honest?

Through calibration on a defined schedule, because resistivity and organic carbon instruments drift and an uncalibrated instrument produces confident wrong readings. Where an instrument triggers an alarm or supports a release decision, its calibration record is part of the evidence for everything it approved, and it is the record most often found lapsed.

What is the commonest device water mistake?

Specifying by grade rather than by requirement, in either direction. Some plants install Water for Injection for a terminally sterilised device and pay for it forever; others rinse precision components in softened municipal water and cannot understand their cleanliness results. Both come from not deriving the specification from what the product needs.

How long does a device water system project take?

Considerably less than a pharmaceutical system, because the qualification burden is lighter, but the design takes longer than expected because the specification has to be reasoned out rather than looked up. The qualification period should still be long enough to see how the system behaves across a few weeks of real production rather than a few days of testing.

Planning a water system in Indiana?

Tell us what your cleaning validation requires of the water. Call 201-450-8280 or use the form below.

Service needed *