Purity Has a Permit: Water Risk in UK Hardware Due Diligence

Water rarely leads a data room. It arrives as a utility bill, a line in a sustainability appendix or a landlord's service-charge schedule. For semiconductor, advanced materials and precision-manufacturing deep tech, that treatment is a mistake. Ultrapure water is not a background utility; it is a process chemical with a specification, a treatment train, a cost curve and a permitting envelope. A company that can make a sample with a small deionised-water skid is not necessarily a company that can run a line. Investors should know which one they are backing before the next capex round.

The process chemical hiding in the utility bill

Ultrapure water is judged by resistivity, total organic carbon, silica, particle counts and microbial control. Those metrics are not decorative. In wafer cleaning, wet etch, chemical-mechanical polishing and precision coating, small excursions can create defects, drift process windows and undermine yield. The water system is therefore part of the process recipe, not part of the building services. It has a design basis, a validation history and a maintenance burden.

The cost profile reinforces the point. Producing ultrapure water requires pre-treatment, reverse osmosis, electrodeionisation, polishing, storage and recirculation. Pumps, membranes, resins and UV units consume power and consumables. Every litre rejected by the treatment train is a litre abstracted, paid for and often discharged. The true cost per cubic metre is rarely visible in a utility invoice, because the capital and maintenance sit elsewhere in the accounts.

Permits are part of the asset base

In the UK, water enters a deep-tech site through several legal doors. A mains supply depends on the water company's capacity and on the site's plumbing and storage. A borehole or river abstraction may require an abstraction licence from the Environment Agency, with conditions that can tighten in stressed catchments. Discharge is a separate matter. Trade effluent consent from the local water company will set limits on volume, pH, suspended solids, metals and chemical oxygen demand; some sites also require environmental permits. These are not administrative afterthoughts. They can determine whether a pilot line can increase throughput, change chemistry or operate at all.

That matters in a UK context where water stress is not uniform but is real. Parts of East Anglia and the Thames Valley have faced serious supply and abstraction pressures. National policy has been moving toward stronger catchment management and more rigorous consenting. A site that looks ordinary on a map may sit in a catchment where new or expanded abstraction is difficult, slow or expensive. Investors do not need to become hydrologists. They do need to ask whether the company's growth plan assumes water rights and discharge capacity that the site does not have.

The gap between a sample and a line

Most early-stage data rooms contain enough evidence to prove that a process works at bench scale. Far fewer contain evidence that water can be supplied, treated and discharged at production scale. The difference is not simply a larger tank. It is a continuous duty cycle, a quality-assurance regime and a permit with headroom.

A sample-making lab can often tolerate batch deliveries, bottled water or a modest polishing loop. A line needs redundancy, monitoring and a water balance that survives peak demand. It needs an effluent strategy for concentrated reject, cleaning chemicals and process residues. It needs a maintenance regime that keeps the loop free of biofilm and particles. It needs a site plan that recognises flood risk, power interruption and single points of failure.

This is where diligence should separate optimism from operational maturity. A company may have a strong materials innovation and a credible customer pipeline, yet still face a twelve-to-eighteen-month delay because its discharge consent does not cover the chemistry of its scale-up. That risk belongs in the valuation, not in a post-investment surprise.

A practical checklist for investors

The first question is the water balance. Ask for average and peak demand by process, the quality specification for each use and the volume of reject or effluent produced. A credible answer will reconcile the process design with the treatment train capacity. A vague answer suggests the water system has not been engineered to the growth plan.

Second, establish the source and the rights. Is the site on mains supply, a borehole, a river or a mixture? Are there abstraction licences, and do their conditions allow the intended expansion? Is the water company's supply capacity confirmed in writing? Does the landlord control any shared infrastructure that could constrain the tenant?

Third, review treatment capacity and redundancy. What is the design output of the ultrapure water plant, and what uptime has it achieved? Are critical components duplicated? How are membranes, resins and UV units maintained, and what consumables budget is assumed? Is quality monitored continuously, and are excursion records retained?

Fourth, diligence the discharge route. Obtain the trade effluent consent and any environmental permit. Check the limits against the scale-up chemistry, not just current operations. Ask about pre-treatment, sludge handling, discharge fees and the likelihood of consent variation. Confirm who bears the cost and liability if limits are exceeded.

Fifth, test resilience. Look for storage capacity, backup power, flood exposure, maintenance access and contingency plans for supply interruption. A single mains feed with no buffer is a different risk from a site with raw-water storage and a validated recovery plan.

Finally, check the contractual and corporate perimeter. Are water rights, permits and infrastructure inside the target company or held by a landlord, a parent or a service provider? Will they transfer on a change of control? Are there environmental liabilities or outstanding consent conditions that could follow the investor?

What good looks like

A well-prepared deep-tech company will not treat water as an ESG footnote. It will have a water mass balance, a permit register, a treatment maintenance plan and a capex line for expansion. It will know the difference between a lab that can make a sample and a line that can run continuously. It will be able to explain how it will negotiate discharge capacity before the first production order, not after.

For UK investors, that evidence is a useful diligence signal. It shows whether management understands the physical and regulatory constraints of hardware scale-up. It also protects against a common valuation error: pricing a company as if its process chemistry were the only scarce asset, when the scarcer asset may be the permit, the treatment capacity and the water itself. Purity has a permit. The data room should show it.