Process Water — Batch and Tank Types: UV System Design
Process-water tank applications look as if lamp placement were trivial — drop a lamp in, set the treatment time, done. In practice, the distribution question becomes decisive as soon as the tank volume grows beyond small feed vessels. UV radiation does not fill a tank evenly; it falls off steeply with distance and is absorbed by the water itself. Lamp count and arrangement are therefore design decisions, not afterthoughts.
Rule of Thumb by Tank Volume
The table below is a practitioner rule of thumb for an initial scoping discussion, not a substitute for a validated reactor design. Required dose, water quality and target organism always drive the final layout.
| Volume | Mixing | Indicative lamp layout |
|---|---|---|
| < 500 L (lab, buffer vessel) | natural convection often sufficient | 1× central immersion lamp |
| 500–2,000 L (feed vessels, small tanks) | light mixing recommended | 1× central, or 2× axially symmetric |
| 2,000–5,000 L (mid-size tanks) | stirrer or recirculation pump needed | 2× distributed (top + middle, or two wall-mounted immersion tubes) |
| > 5,000 L (industrial reservoir, large vessels) | mixing critical | 3–4× distributed across the tank volume (longitudinal axis + cross distribution) |
| > 50,000 L (utility scale, heavy industry) | active recirculation mandatory | multi-lamp array; consider a bypass reactor instead of in-tank treatment |
Why a Single Lamp Fails in Large Tanks
UV irradiance obeys the inverse-square law: intensity drops with the square of the distance from the source. On top of that, water absorbs UV-C exponentially with path length (Beer-Lambert law), governed by the UV transmittance (UVT) of the water.
Consider an illustrative calculated estimate — a lamp delivering 100 W of UV-C output, mounted centrally in a 5 × 5 × 3 m tank (75,000 L):
- Distance from lamp to the nearest wall: roughly 1.5 m radially, 2.5 m vertically.
- For clear water, combining inverse-square falloff with Beer-Lambert absorption, only a small single-digit percentage of the lamp's near-field intensity reaches a wall at that radius.
- The tank corners (roughly 3 m diagonal) receive an order of magnitude less still.
These figures are a back-of-envelope model, not measured data — but the direction is robust: the practical consequence is that the mixing time must be shorter than the treatment time, otherwise some water layers are never adequately irradiated.
Without active mixing, water stratifies by temperature and small density differences, creating "dead" zones that receive little or no UV. This is a classic trap: it is invisible on a datasheet but shows up in a pharmaceutical or food-industry audit. The literature on UV reactor design confirms that mixing moves microbes from non-irradiated into irradiated volumes, lowering the required exposure time and improving fluence uniformity.
Practical Tips for Multi-Lamp Distribution
Arrangement patterns:
- Longitudinal distribution: in elongated tanks (for example collecting tanks in HVAC systems), space lamps along the main axis at a pitch on the order of the tank width.
- Wall-mounted immersion tubes: insert quartz sleeves laterally from the outside rather than letting a lamp float freely. This is maintenance-friendly — the lamp can be changed without draining the tank.
- Floor mounting: in shallow, basin-like tanks, distribute several short lamps across the floor.
- Top entry: suspend several lamps from above, each covering one column of the tank.
What to avoid:
- Relying on a single floor-mounted lamp plus the hope that "thermal convection will mix it" — in large tanks it will not.
- Clustering lamps directly next to each other — this creates a hotspot while the rest of the tank stays untreated.
- For tanks above roughly 50,000 L, plan a bypass reactor instead. Beyond that scale, in-tank UV is frequently uneconomic; a dedicated flow-through reactor with a defined irradiation chamber is easier to validate.
The UV-Transmittance Factor
The lower the UV transmittance of the water (UVT — the percentage of intensity remaining after a defined path length, commonly 10 mm at 254 nm), the shorter the effective UV range and the more critical the multi-lamp arrangement.
The table below gives order-of-magnitude orientation only — the figures are model estimates derived from inverse-square plus Beer-Lambert behaviour, not validated measurements. Effective range also depends on lamp power, geometry and the required dose.
| UVT (10 mm, 254 nm) | Water quality | Effective range, single mid-power lamp |
|---|---|---|
| ~95 % | drinking water, pharma-grade water | roughly 1–2 m radially |
| ~85 % | clear process water | order of ~1 m |
| ~75 % | slightly turbid water | order of decimetres |
| ~60 % | turbid / coloured water | short range — multi-lamp mandatory |
| < 50 % | very turbid water | UV unsuitable; filter as pre-treatment |
As a working guideline, below roughly 70 % UVT a central single-lamp arrangement is generally no longer adequate — even in mid-size tanks (around 2,000 L), several lamps at shorter spacing tend to outperform one high-power unit. Where water quality is poor, pre-filtration to raise UVT is usually the more cost-effective first step before adding lamp power.
Cross-References
- UV Lamp Technology — when to choose low-pressure versus amalgam lamps.
- Material Pitfalls in Practice — quartz sleeve care and ballast positioning.
- UV Economics and ROI — in-tank UV versus bypass-reactor total cost of ownership.
- Drinking Water System Types — related system taxonomy for water applications.
Sources
- Bolton, J. R. & Cotton, C. A., The Ultraviolet Disinfection Handbook (AWWA, 2008) — UVT, dose and reactor design fundamentals for water treatment.
- DIN 19294-1:2020-08, Devices for the disinfection of water by means of ultraviolet radiation — Part 1: Devices with UV low-pressure lamps — type-testing and design requirements; applied here by analogy to process-water tanks.
- Sözen et al., A computational study of the effect of lamp arrangements on the performance of ultraviolet water disinfection reactors, Chemical Engineering Science (ScienceDirect) — lamp arrangement, irradiance distribution and mixing.
- Development and modeling of a novel type of photoreactor for water treatment (NCBI PMC) — reactor geometry, irradiance falloff and dead-zone behaviour.
- In-Situ / UV Alliance application notes on UV transmittance — UVT definition and its effect on UV penetration depth.