Since 1975 granular activated carbon (GAC) has been used by Severn Trent for the removal of microorganic pollutants. The first GAC installation was at Church Wilne WTW, with the primary purpose of removing compounds causing unwanted taste and odour.

Concern about the levels of pesticides in raw water catchments during the 1980s provided the impetus for the installation of GAC plants at other treatment works. Now twelve surface treatment sites use GAC and the process plays an important part of the company’s objective of achieving 99.95% compliance with water quality standards at the tap.

Initial capital expenditure for the installation of a GAC plant can run into millions of pounds and the capital cost of the GAC is a significant part of this. Revenue costs are also high, from £2/Ml to £7/Ml depending on the quality of the raw water to be treated, but they can be reduced by constant monitoring of the GAC bed to predict breakthrough of organic micro-pollutants.

Many types of carbon are available and when placing the original contract for GAC Severn Trent needed to rank the different types. The criteria used were;

  • adsorption characteristics,
  • effects of particle attrition,
  • regeneration losses,
  • purchase cost.

    Iodine number and methylene blue number are both measures of the adsorption capacity of GAC for small and large molecules respectively and Severn Trent specified a minimum adsorption capacity 1,050mg/g for iodine number and 250mg/g for the methylene blue number.

    It was also important attrition losses due to GAC particles rubbing together were low as the GAC would require backwashing at regular intervals. Finally, it was decided that carbon regeneration rather than replacement with virgin carbon was most effective and maximised environmental benefits from recycling the carbon.

    Only two coal-based carbon products met the criteria at the time and two manufacturers were awarded the contract to supply Seven Trent’s 11,000m3 requirements.

    In 1991 Severn Trent set up Grafham Carbons to ensure its GAC regeneration needs were met. The water company gained expertise in the use and handling of GAC but still relied on the original carbon specification, which prompted Grafham Carbons to evaluate a range of different coal-based GACs. The initial work was undertaken at a pilot scale to assess both the carbon’s physico/chemical properties and the effects of regeneration. From these results the whole life cost for the individual products could be compared – a first for the industry.

    Using the test results (see below) combined with typical annual costs the GAC types were ranked in order of performance and a full-scale plant trial was planned. Since GAC is so expensive and performance in terms of adsorption of micro-organics so critical it would have been difficult to justify moving straight to full-scale trials without the reassurance provided by the pilot work.

    The big four

    Of the seven carbons tested four were chosen for the full-scale trials. A number of sites have been supplied with the different types of GAC and their performance is being monitored. The adsorption characteristics with regard to the removal of key organic compounds and the results of the first regeneration trials should provide Severn Trent with sufficient data to aid the choice of GAC purchased both for its own needs and those of Grafham Carbons’ customers.

    Because Severn Trent is looking for value for money in its future carbon requirements new carbon products will be screened prior to acceptance for use by undergoing pilot and possibly full scale-trials. The results will be available free of charge to all Grafham Carbons’ customers in a bid to provide added value on an ongoing basis. Grafham Carbons provides GAC regeneration services throughout the UK.

    Carbon data

    For the pilot plant trials seven coal-based carbons, labelled A to G,

    were evaluated. Each carbon was assessed for; i) losses due to attrition,

    ii) losses on regeneration, iii) change in adsorption capacity prior to

    and following regeneration, iv) evaluation of quality on return to service

    following regeneration.

    The findings were as follows:

    i) Losses due to attrition

    A test rig was built to eveluate the effect of backwashing on batches

    of carbon. The carbon bed was backwashed continually (30% bed expansion)

    for 24 hours which simulated the time for backwashing if the carbon had

    been in service for 18 months. Both sieve analysis and measurement of

    the volume of GAC were carried out before and after backwashing. Each

    carbon was tested in triplicate to ensure repeatability of results. Average

    losses ranged from 0.7% (carbon G) to 3.2% (carbon D). Generally there

    was a slight loss of some of the smaller particles (<600µm).

    ii) Losses due to regeneration

    A test rig furnace, the only one of its type, with a variable speed screw

    feed was used. The furnace incorporated four fully adjustable temperature

    controlled zones. The retention time

    of the GAC in the furnace could be controlled by the speed of rotation

    of the furnace, while the temperature profile throughout the furnace was

    set to mimic that of the full scale furnace at Grafham Carbons. Regeneration

    losses were eveluated over five regeneration cycles on triplicate samples

    for each carbon. The results indicated that the average loss per regeneration

    ranged from 2.7% (carbon F) to 5.2% (carbon B).

    iii) Adsorption capacity as a result of regeneration

    To assess the change in the adsorption capacity of the carbon the iodine

    and methylene blue numbers were measured prior to and following regeneration.

    For the majority of the carbons tested there was an insignificant loss

    of adsorption capacity, the exception being carbon D with a loss of over

    16%.

    iv) Quality on return to service following regeneration

    The ease with which regenerated carbon can be returned to service was

    investigated on pre-loaded carbon samples. Parameters such as pH, chlorine

    demand, aluminium, calcium, sulphate and alkalinity levels were monitored.

    The most notable result was the release of high levels of aluminium for

    carbons A and B.


  • Action inspires action. Stay ahead of the curve with sustainability and energy newsletters from edie

    Subscribe