Effluent: Waste or resource?
Changes in rainfall and increasing demand mean that water has now become too valuable a resource to be used only once. Here Martin Hind of Biwater Treatment looks at some alternatives
Generally, the first response to a water shortage is a clamour to develop new sources. If no natural water sources are available, then desalination becomes attractive as both capital and operating costs continue to fall.
There has been a substantial cost reduction in both capital and operating costs of desalination over the past 20 years. The major component of the operating cost is power. And there have been significant advances in the system design of the desalination plant in addition to major improvements to the membrane and energy recovery devices. Two plants constructed by Biwater are compared in Table 1.
Efficiencies as low as 2.5kWh/m3 are achievable for larger single-stage seawater reverse osmosis (RO) plants. The energy requirement was taken as the total energy required and includes abstraction pumps and pre-treatment. Claims of 2kWh/m3 only consider the power required for desalination.
However, from the Table 2, it is seen that both the capital and operating costs are lower when treating secondary municipal effluent than the desalination of seawater. This comparison was made with ultrafiltration (UF) followed by RO for treatment of the secondary effluent and RO for the seawater desalination.
Sommariva gives alternative operating costs (as seen in Table 3). If production or operating costs are considered rather than life-cycle costs, then the production costs of seawater desalination is twice that of treatment of secondary wastewater.
The technological improvements in the treatment of secondary effluent are perhaps even more dramatic than those in seawater desalination. Table 4 shows a comparison of two water-reuse projects supplied by Biwater.
There are three major reasons for the considerable reduction in production costs. Firstly, the previously used cellulose acetate membranes, which achieved only 96% salt rejection, have been replaced with thin film composite polyamide membranes with a salt rejection of over 99%. The standardised membrane flux of the polyamide membranes is about three times that of the cellulose acetate membrane resulting in the lower feed pressure.
Secondly, there has been significant development of the anti-scalant chemicals required for RO plant which permit operation with a brine Langelier Saturation Index (LSI) up to +2.5, whereas the anti-scalant used in the 1980s was sodium hexametaphosphate which only permitted a brine LSI of +1.0. Consequently, there is no longer a requirement to soften the water and to dose acid. Finally, the extensive traditional pre-treatment can be replaced with micro-filtration or ultrafiltration. Membranes are often considered to be barriers to bacteria and the double membrane approach should result in almost complete removal of bacteria and virus.
Whereas the use of UF and MF membranes has become the accepted pre-treatment for water-reuse projects where there is an existing WwTW, it is expected that membrane bio-reactors (MBR) will be used when there is no existing WwTW. An MBR combines secondary and tertiary treatment for the sewage in a single-stage and will produce an effluent suitable to feed directly to an RO plant. The use of MBRs has become accepted within the UK for municipal effluent treatment where high-quality water, low sludge production and limited land usage is required. (See Table 6.)
Singapore is an island approximately the size of the Isle of Wight with over 4 million inhabitants. Singapore signed a 50-year water supply contract from Malaysia on independence from Malaysia in 1965. This agreement is due to expire in 2015 and the Singapore Authorities are developing their water resources to be less reliant on Malaysia. They have adopted a four-taps concept:
- Tap 1: Continued sourcing of water from Malaysia
- Tap 2: Construction of a 136,000m3/d seawater reverse-osmosis plant at Tuas
- Tap 3: Expansion of water collection and storage through the construction of a dam at the mouth of the Singapore River
- Tap 4: The construction of plants to treat secondary effluent to produce low salinity recovered water. The Singaporeans know this as NEWater
Over abstraction from the aquifer around Los Angeles had resulted in seawater ingress into the aquifer. Secondary effluent has been treated by RO for more than 30 years, and the product water has been injected between the sea and the fresh water aquifer to prevent further sea water ingress. The treatment plant at Water Factory 21, Orange County, California, US, is widely accepted as the centre of excellence for research into effluent treatment.
The salinity of potable water in the UK varies significantly from the low-salinity soft, upland sources in the North and West to the higher salinity hard underground sources in the South and East. The potable water salinity is compared to the salinity in desalinated seawater and recovered water in Table 7. Potable water is treated by industrial users and the degree of softening or salinity reduction is determined by the application.
It is clear that industrial users must have a higher degree of treatment when fed with potable water from the South-east of England than if they were fed recovered water. (See Table 8.) Consequently the value of recovered water to industry is greater in the South and the East of England where the potable water salinity is higher. It is fortunate that this corresponds to the areas of greatest water scarcity. The treatment costs to achieve a water quality suitable for a high-pressure boiler are detailed in Table 9.
Production costs would be reduced slightly if a RO/Ion exchange (IX) system was used for a 400mg/l feed5 when compared to the ion exchange only treatment. (See Table 9.) However, many industrial users have existing ion exchange plants and could not justify the additional capital spend to install an RO plant to reduce the chemical costs in the ion exchange plant.
From Table 3, the cost of treatment of final effluent is £0.16/m3. The saving to industry in ion exchange treatment costs in using recovered water rather than potable can be £0.24/m3. Consequently, the total costs to treat secondary effluent to a standard suitable for a high-pressure boiler feed is £0.36/m3.
It can therefore be argued that recovered water is more valuable to industry and is of a more predictable composition than potable water as the water companies will change the water sources to ensure continuity of supply. However, the low salinity and relative abundance of water in the North and the West gives no economic driver for water recovery in these regions.
Furthermore, the environmental impact of the discharge of spent ion exchange regenerants is significant and this will be greatly reduced as the feed salinity to the ion exchange plant is reduced.
Sommariva argues that naturally occurring water should have a higher value apportioned to it to try to reduce over-abstraction from rivers and aquifers. Water reuse would then become more economically viable if this was applied. This may occur without differential water charges if the Environment Agency (EA) restricts abstraction to maintain river flow rates. The EA is in an unfortunate position because, on the one hand, they may wish to restrict abstraction to maintain the environmental quality but, on the other, they are funded by the abstraction and discharge charges.
The current investment programme (AMP4) may not have the necessary funds agreed for extensive reuse projects, and this may also block reuse projects until the next capital investment programme. Is it acceptable that many areas of the South-east of England will have to endure water restrictions? Alternatively, would those inhabitants accept higher water charges so that they could wash their cars and water their gardens?
The water availability in the UK is 2,724m3/h/d which is marginally higher than Spain at 2,702 m3/h/d. However, 25% of the population of the UK resides in the South-east of England where the water availability falls to 265m3/h/d. For comparison, the water availability in Jordan in 1990 was 327m3/h/d. Therefore we can conclude that the water supply to the South-east of England is very efficient.
While leakage rates may be high, additional water supplies must be developed to ensure continuity of supply as water supply companies gradually repair or replace the ageing infrastructure. The water scarcity problems have been exacerbated by population increases in the South-east and the resources will be stretched further should the planned 500,000 new houses be built. Metering and reduced usage have been proposed, and this will help stabilise demand, but it is doubtful whether this will be sufficient.
Thames Water is planning a desalination plant of 150,000m3/d in the Thames estuary. Although the pre-treatment is more extensive as the feed water quality is poor compared with open seawater, the salinity of the water is around 25% of seawater. Consequently, the power requirement for desalination is reduced significantly. Local politicians have made submissions to the public inquiry to prevent construction until further reductions in leakage have been attained.
New reservoirs have been proposed but this pre-supposes that there is sufficient rainfall to fill them. This is not the case at Bewl Water in Kent, one of the major reservoirs in the area. Local water grids could help spread the resources more evenly. However, there is no area in the South-east of England with abundant water supplies. New resources are therefore required. There are three possible solutions:
- Transport the water from areas outside the South-east
- Construct desalination plants
- Construct recovered water plants for indirect reuse or industrial use
The post-treatment to convert the soft desalinated water to hard London water will also be the same. Therefore, the comparison is purely the power required. Assuming identical pump and motor efficiencies, the pump head for desalination is 629m. Therefore, the total operating cost is identical. However, Northumbrian Water should be recompensed for the sale of water, which would make the operating costs of a desalination plant lower than the transfer of water from Kielder to London. If the owner of the resource was recompensed £0.05/m3, and if the source was at sea level, then distance of the source to London would need to be reduced to 230km to have an equivalent operating cost to desalination.
If the pre- and post-treatment are the same, the capital cost of the pipeline option can be compared with the desalination plant option. Typically a 1m-diameter pipe installed cost is £263/m, giving a total cost of £133M. From Table 2, the capital cost of desalination is £259/m3/day giving a capital cost of £25.9M.
The other issue to consider is that 25 tonnes/d of limestone will have to be abstracted to re-harden the water to London levels. If soft water, whether it is desalinated or natural, is added to the water in the London ring main, it will result in corrosive water. This will result in more leaks as the pipelines corrode. The energy required to convert the limestone to lime and the environmental impact of abstraction of that quantity of limestone should be considered. From the above, it is clear that piping water from Kielder to London is neither cost-effective, when compared with desalination, nor environmentally sustainable.
Recovered water would not be re-hardened as it would not be considered for direct water use, therefore only the capital costs associated with the pre-treatment RO need to be considered. From the discussions above it was shown that the capital cost for the treatment of final effluent is 34% less and the running costs are half that of seawater desalination. This would appear to be the most sustainable approach. The recovered water could then be used for industry, where the recovered water would be more valuable than potable water, or it could be used for indirect water reuse either through reservoir augmentation or aquifer re-charge. Experience at Langford WTW in Essex and Suffolk water suggests that public perception will not permit the direct reuse of water.
There is extensive ongoing research regarding the discharge of sewage into rivers and the effects of endocrine disrupting chemicals (EDC) on the fish population. The majority of male roach are classed as inter-sex. It should be noted that RO membranes will remove around 99% of EDC from the water. Therefore, it can be argued that recovered water can be of a higher quality than the potable water that is currently supplied. While there is no known risk to human health, it is likely that there will be public concern about this when it becomes more widely known.
To put this approach into context, there has been unplanned water reuse for decades where treated sewage is discharged into a river, which is then used as a water source downstream. During certain periods, the flow in some rivers is largely sustained by the flows of discharged effluent. There are three major reuse possibilities in the South-east of England:
- Industrial use of recovered secondary effluent: Anglian Water installed a water recovery plant at Flag Fen WwTW to supply low-salinity water to the local power station, resulting in reduced ion exchange treatment costs for the boiler feed water. In this instance, the power station was located close to the WwTW, which may not be the case in other areas and a local network of recovered water would be required to supply water to the industrial users.
- Local MBR: If the WwTW is some distance from the industrial user, it may be more attractive to take untreated sewage as a feed source and treat using a locally based MBR and RO. The concept of sewer mining may become possible.
- Aquifer re-charge or reservoir augmentation: While it is believed that public perception will not accept direct reuse, the concept of indirect reuse is already well established. Secondary effluent could be treated to produce high-quality water that is then discharged to streams that feed reservoirs or pumped into aquifers. The water will then be mixed with natural waters and would be treated by a conventional WTW.
Designated projects have been agreed between the Water Companies and Ofwat, and these have been fixed until 2010. Water charges are linked to this agreed capital investment plan. If major water reuse schemes and the construction of local distribution networks are to be considered, then water charges will have to be increased. Will water users accept these increased charges when they consider the high water losses that have been publicised in certain areas?
There will be negative public perception of water reuse projects for indirect potable use such as reservoir augmentation and aquifer re-charge. The Water Companies will probably not wish to point out that this has been happening for years as water is abstracted from rivers for treatment downstream of sewage discharges. However there may be pressure to stop this practice if any link is established between the decline of human fertility and the consumption of water contaminated with EDC.
Will the three water regulatory bodies, Ofwat, DWI and EA, have a role in the recommendation or prevention of the indirect use of recovered water?
The supply of water to the South-east of England is very efficient and, although savings could result from the ageing infrastructure being replaced, additional water sources are needed. The construction of more reservoirs is not necessarily the answer as there may not be sufficient rainfall to fill them and the time taken to obtain planning approval and construct them can be significant. The value of the land lost when flooded must also be considered in the economic evaluation. A national water grid, or any long pipeline to take water from the North of England or even Scotland, is as environmentally sustainable as desalination but carries a far higher investment cost. This approach is therefore not considered viable.
Seawater desalination can be viable for coastal users, in particular where the desalination plant feeds a discrete delivery network and the amount of lime required to give the desalinated water a similar composition to the existing potable water can be reduced. It has been shown that not only does the technology exist for the treatment of sewage to produce water that is of a higher value than potable water but this technology has been applied in other countries around the world. Furthermore, the value of the recovered water to some industrial suppliers who treat the water further for boiler feed or other process applications is higher than that of potable water.
Consequently effluent water can be considered as a valuable resource. The development of recovered water plants may be a local industrial initiative but it is more probable that a government initiative is required, particularly if they wish to pursue the construction of thousands of new homes in the water-stressed areas of the South-east of England.