An in pipe system rebuild.

This document has been edited to remove any confidential information.


The Installation.

We are only concerned with the UV treatment plant.

The system consists of a series of three modules situated in a sunken enclosure with the supply and control boxes situated remotely above ground.clip_image002_002

Each module consists of a stainless steel cylinder mounted horizontally on a substantial frame. There are entry and exit pipes connected to either end.

The stainless steel cylinders are in two parts, an outer shell and an inner core through which the effluent flows.

clip_image004_002-1clip_image006_002clip_image006_003clip_image018The system works by irradiating the effluent as it is passed through the hellictical core of the module

Arranged horizontally around the full circumference of the inner core is an array of 18 lamps, which produce a high level of ultra violet light in the required spectrum.

The lamps are enclosed in open-ended quartz sleeves, which maintain separation from the effluent whilst permitting the maximum transmission of UV light to the effluent.

There is a water spray system fitted to each module, this is designed to prevent the build up of dirt on the sleeves thereby maintaining the proper level of UV transmission.

The control and distribution system for this is contained within the enclosures which can be seen mounted on the trunking immediately adjacent the modules.

The lamps are cooled by a fan mounted in the end panel of the cylinder.

There are two types of lamps, which can be used in this system both are manufactured in the US.

The manufacturer recommends a lamp produced by Lighttech as it is supposed to be more easily struck up than the alternative produced by Voltarc.

There is no measurable difference in the performance of the two lamp types and either should give perfectly satisfactory performance.

The Voltarc lamps do have a significant advantage in that

they do not require preheating prior to start-up and are more readily available than the alternative.clip_image010

Power to the lamps is supplied by a quite complex wiring loom, which surrounds the outer cylinder.

These looms extend above ground into the control and reporting cabinets , the cable clusters are contained in flexible conduits.clip_image012






Control and reporting Cabinets.

There are three cabinets one for each module.

clip_image018 clip_image020

UV intensity-measuring system.

Each of the modules has a UV intensity-measuring device, which reports to the control panel boxes.

These have been the subject of some modification in order to improve their performance and we believe the performance is now satisfactory.

This system is independent of the supply and reporting system, which serves the UV lamps, it need not be disturbed by the works proposed.

The Problem.

There is an excessive rate of failure of the ballasts leading to an under performance of the system.

The maintenance burden is similarly much greater than should be the case.

It has not been possible to identify a single cause for the failures of the ballast cards.

We have asked our Electronics engineers to investigate the problem but they have been unable to offer a solution.

They have tried exhaustive tests and have even replicated the ballast card but it is inherently unstable and the maximum operating life we have been able to produce is 150 hours.

We understand that the manufacturer has similarly been unable to provide a more reliable version.clip_image022

The Solution.

As the existing ballast cards appear to be incapable of improvement an alternative product is required.

We have identified a suitable generic ballast produced by a reputable manufacturer.

This manufacturer has a good track record and South West Water uses a considerable quantity of the Company’s products at other locations.

The product is readily available in quantity and we have been assured that it will be manufactured for the foreseeable future and is widely used in the industry.

Future availability should not therefore be a problem.

There is a major difference in the operation of these replacement ballasts; they need to be situated within 4 metres of the lamps.

The existing ballasts are located some 6 to 8 metres from the lamps and therefore the existing location points cannot be used.

The replacement ballasts are also substantially larger than those currently in use.

The Trial.

Mr John Taylor installed two of the ballasts we supplied in the water spray control panel adjacent to the UV modules; this was of course a temporary installation and the ballasts were not fixed in any way.

A power supply was connected and a connection was made to the lamps.

The temporary installation did not address the question of reporting of the lamp’s performance as should be the case in the permanent installation and so it required manual observation to confirm its performance.

This was of necessity a very basic installation, the ballast were not cooled in any way

The performance was excellent in that no action was needed over a period of some two months the ballasts and lamps performed exactly as they should.

The trial must therefore be considered a complete success.

The project.

Project is best described by following the sequence of actions required.

Fortunately it is possible to work   with the modules on an individual basis therefore no more than one module need be “offline” at any time.

Each module will take approximately 5 working days to modify it would however be prudent to allow 7 working days for the first module in case unforeseen difficulties arise.

In order to ensure that the system does not fail consent during the period of the works it has agreed that temporary works will be carried out to ensure that all tubes are capable of performing their disinfection function despite the module being offline.

This will be achieved in the following way.

All lamps and tubes in all three modules will be replaced prior commencement of the works, this is to ensure maximum performance of the system.

Prior to any disconnection or isolation, of the module selected to be modified, the new ballasts will be connected to the lamps and a temporary power supply connected to allow the tubes to perform their disinfection function although they will not be capable of reporting the local operator interface or the SCADA system.

The lamps may be monitored in a visual manner until such times as the revised reporting circuits are in place.

In this way we can ensure no disinfection capacity is lost during the conversion process only the ability to monitor performance remotely.

The programme of works following the above is as follows.

Isolate control panel to one module.

Remove existing ballast cards and set aside.

Remove existing wiring to cards and discard.

Remove redundant ballast mounting frames and discard.

Fit blanking plates to door mountings.

Rebuild looms and reroute existing power feeds and control wiring looms   into new terminal rails (see below.)

clip_image002_003                   clip_image004_003

Remove existing wiring power supply cables and control wiring from flexible conduits and discard.

Maintain separation of UV reporting system at all times.

Install new cable runs to module location (72 number) through existing flexible conduit if possible, if not possible then renew.

Install new (Rittal) ballast enclosure cabinet to the rear of the existing trunking and support system,

Enclosures to be 600mm x 1200mm x 300mm.


Divert existing conduits and wiring loom from modules to the new enclosure.

Divert power supply to existing cooling fan to new enclosure.

Install 9 number Voltarc ballasts to prepared panel.

Install 3 number purpose built distribution and reporting boards onto prepared back panel.

These boards will report to the PLC   lamp performance but only to the extent of whether or not they are functioning properly the UV intensity is measured by a separate system.


Install cooling system to new ballast enclosure, we have calculated that a simple fan system will suffice to properly cool the ballasts,

The ballasts are fully enclosed and IP65 rated.

An air conditioning system can be fitted to the top surface of the enclosure if this is a preferred option.

Install new wiring loom to the module flange from the new enclosure.

Install new power feed to the lamp-cooling fan.

New UV lamps to be fitted to the cylinder (18 number) this is due to the fact that Voltarc lamps will work more efficiently with the Voltarc ballasts and will not require preheating as do the Lighttech lamps.

Test entire ” below ground” installation for continuity and insulation resistance using temporary power supply to ensure installation is satisfactory prior to connection to the main control panel terminal rails and the PLC.

Measure and calibrate reporting signals prior to connection to main control panel and PLC.

When tested satisfactorily connect to the main terminal rails and the PLC.

clip_image012_000  clip_image014_000         clip_image016_000          

Install new reporting links to the PLC

These links will ensure that the existing LED reporting indications will continue to operate as before allowing immediate visual inspection   at the control cabinet as well as on the SCADA system without the need to open the enclosure.


Systematic commissioning will be required to ensure that the correct signals are being given and received by the PLC and that there is complete compatibility with the SCADA system.


This project has been prepared in consultation with the personnel responsible for the day to day maintenance of the system and we are confident that it is both practical and implementable with the minimum of disruption

We are confident that when completed the system will be efficient, the reporting systems will be as before, the maintenance burden will be considerably reduced and above all the system will be reliable.

Lamp Life and rated output.

Hours output















































Since preparing the above a further question has been raised regarding the total UV out put of the system.

Apparently it has been calculated that there may be a ” shortfall” in the applied UV dose of approximately 10% in the system under certain circumstances.

These calculations are carried out assuming the UV lamps are at the end of their rated life.

It can be seen that from the attached graph and the raw data schedule that by replacing the lamps at 6000 hours rather than 9000 hours the UV level will always remain at least 10% above their minimum output.


This is the most cost effective means of overcoming the potential problem.

The cost of doing this is approximately £ 855.00 per annum.

In practice we find that the manufacturers tend to be very conservative with the output figures and we invariably find that performance is well above the minimum they predict at any given point on the graphs.

This is particularly evident when the ballasts and lamps are ” matched products” i.e. specifically designed to operate together as in this case.