Improving storage facility quality

In 1997, following a
reversal in the improved trend in compliance seen in previous years, a renewed commitment was made by Yorkshire Water to improve the quality of storage facilities. No single initiative has been responsible for the steady improvement seen since then – success has been achieved by a multi-faceted approach across the company.

This improvement has been demonstrated by a ten-fold reduction in coliform detections over a period of six years. Storage of treated water is provided in 414 tanks and towers of various construction types ranging from 148,000m3 underground reinforced concrete structures to a 0.01,000m3 stainless steel tank. It was recognised that any actions taken to bring about a company-wide improvement in the compliance of storage facilities needed a co-ordinated approach by all operational areas.

A service reservoir group was established with representatives from regional engineering, asset owners and
managers, water quality and capital development. The group identified broad areas of concern, some easily remedied (for example, by improvement of sampling facilities) while others needed major engineering improvements requiring significant capital investment.

Storage was identified as a group of key assets which, because of their widespread and often isolated distribution, may have not received the attention and investment their role justified. Improvement can only be based on information and knowledge, so detailed evaluations of each structure was seen as a necessity to longer term improvement in performance and water quality.

The quality of water leaving the company’s WTWs has improved significantly both chemically and bacteriologically. The specific requirements of the Water Supply (Water Quality) Regulations (Anon. 2000) effectively directed water utilities to optimise treatment such that turbidities leaving works were minimal. This has resulted in less turbidity and suspended material being deposited in distribution mains, as well as allowing disinfectant residuals to penetrate much further into the system. In parallel with these treatment improvements there has been a major capital programme of distribution system renovation and improvement. These actions are believed to have significantly improved the quality of water delivered to storage reservoirs and made a major contribution to the improvements discussed here.

The integrity and suitability of the storage structure is fundamental to the maintenance of water quality. In any water utility, storage facilities will differ greatly in design, age and present day suitability. Facilities built 100 years ago, by today’s standards may well be inappropriate. Accordingly, it has to be decided whether or not abandonment or extensive renovation is an option – replacement costs for very large, older storage may be prohibitive and difficult to
justify. For extensive urban supplies, storage is almost invariably large multi-compartment reinforced concrete structures but for rural supplies, glass reinforced plastic (GRP) has served as an alternative, although in recent years buried stainless steel has proved successful for small, new-build tanks.

Older, problematic steel tanks have been phased out. The elevated temperatures created in these tanks, especially during summer, was found to have a negative impact on water quality, including the loss of disinfectant residual, increase in colony counts and a tendency to give rise to taste and odour problems where tank stagnation was common. Deterioration of the steel structure through corrosion was also found to be a problem, necessitating regular and expensive maintenance. With Yorkshire Water extensively serving rural areas, small, easy-to-construct GRP tanks were once installed extensively. While quick and cheap to install, the nature of their construction has led to problems.

Seasonal temperature variations were found to have a deleterious effect on the integrity of the joints on certain designs. Internal inspection revealed signs of joint breakdown and ingress, with roof flooding during inspection invariably confirming this. Repairs to GRP joints have generally been found to be unsuccessful. Down-time for repairs and the low success rate of such repairs prompted a look for other options.

A recent initiative to overcome problems on smaller storage tanks has been the construction of an encapsulating ‘barn’ over the tank. This has been applied to both steel and GRP tanks, and provides protection from the weather and contamination risks while minimising temperature fluctuations within the tank.
An additional benefit is that a barn in a rural area is much more aesthetically acceptable than the exposed tank. There is also an added element of security in that the tank is hidden from view. A structured programme of inspection, cleaning and maintenance of all reservoirs means all storage facilities are programmed for inspection and cleaning at least once every five years.

This approach not only keeps storage clean but, possibly more importantly, allows a regular opportunity to check for deterioration, faults or potential problems. Programming is based on a points system – points are linked to such considerations as reservoir structure and known faults, age, turnover time, water source type, time since last inspection and water quality indicators. A benefit of the ranking system is the points total can be reduced by actions other than cleaning.

Attention to a known fault or improved turnover, for example, may push the need for cleaning and inspection forward, thereby reducing operating costs. Conversely, where reservoirs cannot be taken out, as in the case of single-compartment storage, the accumulation of points draws attention to that asset and may help proactive management. Integrity testing, by roof flooding during internal inspections, has proved invaluable in detecting ingress. This is now standard practise for all inspections of ground-level storage.

Flood testing and ingress detection generate actions, which may range from a minor seal repair to a major capital scheme. Ingress in the vicinity of the reservoir outlet stands a greater chance of creating a coliform detection at the outlet main sample point than ingress elsewhere in the reservoir, which might remain undetected due to the effect of residual disinfection within the reservoir.

This approach has been found to highlight minor problems at an early stage, avoiding potential regulatory non-compliance and expensive remedial action at a later stage. Roof integrity, especially on underground storage, has proved to be the single most problematic area in terms of direct water quality deterioration. Ingress via cracks in the roof, around the bases of ventilation stacks and access points, via faulty drainage channels or between roof and wall sofit joints are all linked to coliform contamination.

Traditionally in the UK, underground storage has been covered with soil and grass. However, soil is a medium full of microbiological diversity, while grass may attract small mammals. In Yorkshire, active steps have been taken to remove soil and replace it with suitable grade gravel of a size that prevents vegetation growth. While maintaining an insulating barrier, drainage is significantly improved and the microbiological habitat is replaced by an inert substitute.

Soil removal also reduces maintenance in terms of grass cutting, which can be significant if large numbers of reservoirs are involved. There may be local planning issues that mean soil and grass removal is challenged – an artificial alternative may be an option.

When soil is removed, roof membranes are usually replaced. Damaged, loose-laid membranes have been found to entrap water. ‘Tracking’ under such membranes allows ingress of septic water into reservoirs with faults on the roof. Yorkshire has found spray-applied and adhered sheet membranes to be much more effective. Damage to such membranes remains localised and the integrity of the membrane as a whole is uncompromised. Ventilation stacks on the roof of storage reservoirs, especially underground storage covered with soil, are prone to ingress around the base as they deteriorate with age. The removal and replacement of such structures has improved and protected water quality. This, together with the renovation and replacement of all access covers with high-security vented covers, is programmed for storage across the company. Improved quality covers also minimise the chances of access by insects and other small animal life, which seemed to present a challenge to loose or ill-fitting GRP covers. Sometimes problems are caused by oversights or lack of attention to detail.

New concrete reservoirs would not normally be expected to fail bacteriologically due to ingress. However, an example of an open channel in an upstand left over from construction work was allowing direct access to the water. This was only discovered during roof flooding undertaken as part of the clean and inspection process. Access points for telemetry cabling can also allow water, and even small animals, into the reservoir.

This could be caused by personnel who may not appreciate how reservoir integrity could be compromised by their work. Raising the awareness of such problems across Yorkshire Water has brought about a significant reduction in problems of this nature. On certain extended systems with long residence times, chlorine residuals invariably decline to very low levels.

The preferred choices to combat this are booster chlorination or chloramination. Chlorination has been used to maintain a disinfection residual across the distribution system, including storage, but the downside has been adverse customer reaction in terms of chlorinous taste and odour and, on certain systems, a tendency for elevated trihalomethane (THM) formation. The adoption of chloramination on selected systems has proved very successful to date in maintaining a residual within long systems, while at the same time maintaining the best of water quality.
The benefits have been clearly illustrated by a reduction in chlorine complaints from customers on a system previously boosted by a series of chlorination units. It has also resulted in a combined chlorine residual reaching storage reservoirs that previously had little or no residual.

Introduction of a single point of chloramination on an extended system allowed a series of booster chlorination units to be discontinued, giving both maintenance and operational savings. It is important every water quality exception is thoroughly investigated. This is essential for providing information on the failed sample and how it occurred. Identifying the cause of exceptions will help identify other sites that may give rise to the same problem.

Such a proactive approach has contributed significantly to improvement in compliance. An initial assessment of water quality being delivered to the storage in question is made wherever possible. This may indicate upstream problems and avoid the need for expensive investigations of the implicated reservoir or tower.
An assessment of the sampling facility should also be made. Are there problems with the sample tap, sample line or in the case of a pumped sample, the quality and operation of the pump? Similarly, attention should be given to the appropriate flushing time at the tap prior to taking the sample, so all samples are representative of the water held in storage. The use of large-volume samples have been found useful to identify possible low-level contamination, which may be difficult to detect by standard 100ml analysis. Similarly, ‘first draw’ samples at the sample tap are useful in providing information about any activity within the sample line, which may be linked to the coliform detection. Over the last six years the investment of time, effort and capital has contributed to a ten-fold reduction in bacteriological non-compliance as measured by coliform detections at storage reservoirs and towers across the Yorkshire area. Total detections for a calendar year fell from 131 in 1997 to ten in 2003, which equate to 99.37% and 99.95% compliance respectively. Non-compliance has been presented as cumulative detections to highlight the seasonal impact as demonstrated by the increase in detections at the onset of summer. While little can be done to influence water temperature, improved treatment and hence a reduction in available nutrients in supply will be reflected in less after-growth within the distribution system, which in turn means fewer coliform detections at reservoirs. The relationship between storage facility size and frequency of coliform detection was reviewed but, while there were some indication that smaller facilities yielded more detections, the level was not significant.

It would seem the quality of the water supplied from the WTW is more critical than the capacity of the storage facility. Conversely, in the past, coliform detection at storage facilities could be directly attributable to inadequate and ineffective treatment. Improvement in the quality of water held in storage has been shown to be achieved by a combination of actions and initiatives. No single issue will eliminate deterioration of water during storage. However, the treatment process holds the key to achieving the best possible water quality supplied for storage.
A product treated to comply with all water quality standards should subsequently be delivered to storage across the system by a well-maintained and well-managed distribution system. Maintenance of a protective disinfectant residual across the system and within storage is to be encouraged, but not to the point where disinfection byproducts are formed or consumer complaints are generated.
The design, material and suitability of the structure has been shown to be an important consideration in attaining and maintaining the best water quality during storage. The regular cleaning and inspection of all storage reservoirs has been found to be particularly beneficial, with remedial work carried out as soon as possible. Improved awareness and communication can lead directly to improved storage protection. A single co-ordinated and dedicated team working exclusively on all storage facilities across the company has been key to the progress made in recent years.

Further improvement can only come from a continuation of those initiatives – water is a liquid foodstuff and the quality of storage we provide for it should reflect that

l The authors would also like to acknowledge the major contributions made by Paul Scargill and the reservoir cleaning and maintenance team to the successes outlined in this paper. This article is based on a paper presented at Storage 2004.
References:
The Water Supply (Water Quality) Regulations 2000. Statutory Instrument 2000 No 3184. HMSO, London. Anon. 2000.
Unpublished data. AJ Wetherill AJ and JG O’Neill.