Login 
Dr Clifford Jones investigates the oil production by product of produced water, and outlines the various means employed to remove contaminants from it, thereby rendering it usable again for a variety of applications.
Background
Water exits a well with the oil, and separation is fairly straightforward in view of the immiscibility of the two. Obviously, however, the water so separated is contaminated with oil and cannot be released into the environment without prior cleansing.
An oil well once drilled and completed – in the precise sense of that term in petroleum engineering – has a considerable lifetime, and the water-to-oil ratio will not be constant over the period of usage of the well. The amount of produced water relative to oil can become so large as to make the well non-viable, in which case the well is said to be ‘watered out’. At a particular well the produced water tends to be less saline than seawater, although exceptions to this rule are by no means unknown.
It is stated by NETL (the United States’ National Energy Technology Laboratory)1 that worldwide 2.1 to 3.1 barrels of water are ‘produced’ for every barrel of oil yielded. The daily production of oil internationally is 80 million barrels, so taking the mean of the range above gives a value ≈ 200 million barrels of produced water per day, equivalent to 75 billion barrels annually.
There are wells in the United States operating at 50 barrels of water per barrel of oil, and the stage at which a well becomes watered out as described above is to a large degree an arbitrary decision.
Analysis of produced water
That at any oilfield the composition of produced water should be known is obviously important and consequently analytical procedures have been developed. Hydrocarbon molecules fluoresce when exposed to particular wavelengths of radiation; that is, they emit at a higher wavelength from that absorbed. Clearly this can be the basis of measurement of hydrocarbon amounts, and the table below begins with details of two commercially available analytical instruments for produced water that work along these lines.
GC-FID (next row) stands for ‘gas chromatography-flame ionisation detector’, a most widely used technique in analytical chemistry. The basis is that the produced water is admitted to a chromatography column (either ‘packed’ or ‘capillary’) whence it elutes to a flame ionisation detector which gives a quantitative response.
If in general application of GC-FID one admits a coal tar to such a column only a small amount of it will ever ‘elute’ and the remainder will lodge in the first centimetre or so of the column. Obviously this has to be avoided in produced water analysis, and the ISO standard for this3 specifies that hydrocarbons in the range C10 to C40 will be eluted from an ‘apolar capillary column’ which, clearly, overcomes the retention difficulties with standard columns.
C40 would represent the wax component of crude oil, while C10 is roughly in the gasoline range. Excluded by the method then are the lightest components and the heaviest; asphaltenes, the heaviest components of crude oil, can have molecular weights a factor of two or more in excess of those of waxes. The points of detail in which the Ospar (next row of the table) method differs from ISO 9377-2 are minor, one being that the carbon number range goes down to C7.
A reader will immediately, on reading of the infrared methods in the following row, have thought of the difficulty whereby the solvent used in extraction will interfere with the infrared measurements. This is an important point, and for many years was overcome by use of Freon 113, chemical formula Cl2FC-CClF2, which can be relied upon not to absorb in the wavelength range that the oil would.
This was once widely done and was the basis of a method for produced water analysis issued by the American Society for Testing and Materials (ASTM). Freon is of course a chlorofluorocarbon (CFC), and use of these was restricted after the Montreal Protocol in 1995 because of the effect of fluorine on the ozone layer.
Developers were therefore faced with the challenge of finding solvents transparent in the infrared wavelength range to which the oil responds. Alternatively a solvent not having such transparency can be used if there is total evaporation of it before the hydrocarbon of interest – the oil in the produced water – is totally evaporated prior to exposure to the infrared beam.
This has been successfully carried out in countries including Norway with pentane as the solvent; n-hexane has also been so used. Even so, the consensus seems to be that ‘transparent’ solvents are better and considerable development work into these has been reported. The Montreal Protocol classified fluoro compounds according to ‘Ozone Depletion Potential’ (ODP) and did not ban fluoro compounds having a low rating on the ODP scale. Accordingly dichloropentafluorpropane, having a lower ODP than Freon 113, has been considered as a solvent for produced water analysis and is being evaluated in round-robin tests with the involvement of the ASTM.
Treatment of produced water
A great deal of produced water is re-injected into the oil fields from which it came. One of the most widely discussed examples of this (e.g.7), probably on account of its scale, is Prudhoe Bay in Alaska, where the oil yield is about 0.3 million barrels per day from 836 wells.
An illustration of the field is below. The produced water at Prudhoe Bay8 is about three times that, consistent with the international figures given earlier.
The produced water at Prudhoe Bay is re-injected via water injection wells, of which there are 183. There are also wells for gas re-injection.
In general, produced water re-injection is primarily a means of removal of the produced water. Water injection is a means of enhanced oil recovery (EOR), but although it is possible for the water for such a purpose to be the produced water, seawater is also used.
Produced water is sometimes not the most suitable medium for enhanced oil recovery as, irrespective of its salinity, it can contain contaminants absent from seawater. These include microorganisms and radioactive substances.
At Prudhoe Bay both produced water and seawater have been used in EOR. This involved movement of the produced water from the production wells to the water injection wells, and the effect between the numerous water injection wells, as noted above, was to ‘flood’ the field providing for very effective EOR.
At an offshore field release of produced water in the sea is an obvious approach, but not until the produced water has been suitably treated. There are a miscellany of processes available for this, and membranes and electrolysis among other approaches and principles have been applied in devices made available by the oilfield services companies.
Cyclones, osmosis and centrifuges have also featured in such devices. Often, more than one of these techniques will be applied to a particular source of produced water, and produced water having in this way been brought to a very high standard of freedom from contaminants is said to be ‘polished’.
That the abundance of produced water, its composition and opportunities for its re-injection vary widely internationally has already been emphasised and the cost of treating it depends on such factors. A 2012 report by the Society of Petroleum Engineers gives a range of 5 cents to $3.40 as the cost of treating a barrel of produced water across a number of selected fields worldwide.
Without examining the figures further we can obviously conclude that the lower limit is for simple re-injection at source, and that the upper limit applies where there has been heavy capitalisation in treatment plant of the sort described earlier.
The simplest and most obvious means of ‘treating’ produced water has been deferred for discussion until the end of this section of the article, partly because it achieves only separation of the oil without regard to the other contaminants. The method referred to is gravity settlement, and it might be used alone or it might be the first step in a ‘polishing’ process.
Crude oils are slightly less dense than water, so once the produced water is in a settled state only time is needed for separation. That the water is settled is important – if it is not, centrifugal action is needed.
The behaviour often observed in gravity settlement is that small droplets of oil coalesce to form larger ones which migrate to the water surface where they form a film. Once the settlement device has done its job the water layer is analysed, as if there is to be no further treatment the oil remaining in that will be discharged.
If the produced water initially came from an offshore field such discharge will be back into the sea and limits in the range about 10 ppm (parts per million) to 40 ppm will apply to the discharged water.
Further treatment is more likely to be required if the produced water is from an onshore field, and it is not unknown for thoroughly treated water to be passed along for irrigation use.
The interesting point has been made that produced water can in two senses be seen as yielding ‘positive revenue’. The practise just described is one, but it is rare and probably too insignificant to feature in economic calculations widely. The other sense in which it is an asset rather than a difficulty is that, as we have seen, produced water sometimes provides an alternative to seawater as an agent in enhanced oil recovery.
This section of the article concludes with a caution to an inexperienced well engineer not to expect ‘a barrel out for a barrel in’ when water injection is applied to oil fields. In fact, it is common for about 0.75 barrels of oil to be realised by a barrel of injected water. This is because of uptake of some of the water by the geological formation, and clay in particular stimulates such uptake. A formal way of expressing this is that the continuity equation in physics cannot be applied because the mass flow rate is not constant.
Further discussion
Another perspective on produced water is the following: it is easily shown* that a barrel of refined oil products burned leads approximately to a barrel of water in the environment. It is true that not all of the 80 million barrels of oil produced in a day are put to fuel use. Some are used to make petrochemicals, but a great deal of these go into short lifetime products such as packaging, which are soon incinerated.
For the purposes of the present discussion this is as if they were incinerated at source, and we can state that roughly 80 million barrels per day of water enter the atmosphere from combustion of the products of crude oil. Add to that the produced water at the wells, and the water output from the entire oil business – from production to burning of the refined products – is of the order of 300 million barrels per day, or about 50 million cubic metres per day.
Conclusion
‘Water, water every where…’ Those words from Coleridge’s The Rime of the Ancient Mariner might reasonably be applied to the oil industry. They were written in the closing years of the Eighteenth Century, so predated by several decades the boom in coal production to meet the needs of industrialisation. They might in fact have been applied to that, as flooding of coal mines was prevalent in the early years of coal mining on a large scale.
* Detailed calculation available from the author on request.
Published: 27th Nov 2013 in AWE International
