A typical fracture stimulation stage requires approximately 1 million gallons of water. Depending on the specific site, you would expect between 25 - 75% of this fluid to return to the surface in the days following the frack. The fluid that comes back is called "flowback".
The injected fluid may contain about 1% chemical additive. The two principal additives tend to be friction reducers (which reduce the energy needed to pump the water) like polyacrimide (found in many cosmetic products) and thickeners (to help the fluid carry proppant) like guar gum (found in many food products). Cuadrilla have listed their ingredients here.
In addition, during its brief sojourn underground, the injected fluid can pick up additional material from the shale rock, including naturally occurring radioactive material (so-called NORM) and other minerals. Therefore, these fluids need to be treated before they can be safely returned to the water system.
The need to dispose of produced water is not a new problem for the oil industry. In most conventional hydrocarbon reservoirs, there is a certain amount of water trapped along with the oil. As the oil is produced, so is the water. As the field gets older, and most of the oil is gone, more and more water is produced alongside the remaining oil, and the "water cut" (the percentage of water produced alongside the oil) can be as high as 90% (i.e. 90% of the fluid produced from the reservoir is water, not oil).
As is the case with water injected for fracking, this water will have been in contact with the hydrocarbon reservoir, in this case for millions of years, rather than a few days. Therefore it may be highly saline, and it may contain NORM. This water, when produced, has to be dealt with.
A recurring theme in many of my posts is that what is proposed for UK shale extraction does not differ hugely from conventional operations, and also that we have a successful onshore industry with an excellent track record of minimising its environmental impact.
To make this point again, this time with respect to dealing with water produced from oil and gas wells, the table below lists the volumes of water (in barrels) produced from onshore UK oil and gas fields in 2012.
The larger UK offshore industry has to deal with an eye-watering (excuse the pun) 1.5 billion barrels of produced water every year!
A typical hydraulic fracture stimulation will use approximately 1 - 5,000 cubic meters (200,000 - 1,000,000 gallons) of water, or 6,000 - 30,000 barrels. Assuming that 50% returns to surface, that's 3,000 - 15,000 barrels of water to be processed for each frack stage.
This, again, is where comparisons with the conventional industry become pertinent - the current UK onshore industry has to deal with 73,000,000 barrels of produced water every year. It would take thousands of frack stages every year just to to match that rate. As an unnamed onshore operator commented to me recently (my emphases): "most people don't realise that all oil and gas wells produce water, more water than oil, and we have been dealing with it since way before shale gas. It is a lot of water. Can't blame the UK water on fracs!".
I'm not a chemist, so I'm not really up to speed with the details of how produced water is processed. I had an anonymous commenter on this post who seemed to really know his stuff. However, the best place to find out would be from the guys who are doing it, such as Lee Petts from Remsol, who has blogged about the issue here.