It is commonplace in drilling for oil and gas, to use fluids such as brines as well servicing fluids. In general, such fluids have desirably a high density. In particular, highly concentrated alkali metal salts of carboxylic acids have been developed for use in oil and gas well drilling and completion operations. Such a fluid may be based upon cesium and potassium salts such as formates, which can provide brines of specific gravity values of 1.6 to 2.3, depending upon solution concentrations.

During use of such brines, they may be contaminated with sodium chloride from sea water or from water or salt entrained within the rock and mud during application in the well bore. Chloride ions can cause severe corrosion of steel pipework and additionally may be incompatible with the rock matrix, causing damage to or near the well bore area. Such problems-. may result in users of such brines refusing to re-use chloride-contaminated brines.

A charge of cesium formate solution for use as an oil field brine may cost in the order of $1 OM, so there is considerable need to improve the prospects for recycling such a brine.

Further, if a diluted brine is returned to high density by evaporation of water, any chloride concentration will increase by this process, thereby rendering the brine less suitable for re-use.

Accordingly, the present invention can be applied to removal of chloride from brine both caused by contamination in use or caused by a concentration process such as evaporation, and throughout this description and claims the term"brine"is to be understood as including brines which have been concentrated.

The present invention provides a method of treating an alkali metal carboxylate salt brine contaminated with chloride ion, comprising mixing such contaminated brine with a

solution of a silver salt, especially silver nitrate, causing silver chloride to be formed and separating the silver chloride from the residual brine.

The invention may also be expressed as a method of use of an alkali metal carboxylate salt brine, comprising the recovery of used brine contaminated with chloride ion, treating the recovered brine with a solution of a silver salt, especially silver nitrate, causing silver chloride to be formed and separating the silver chloride from the brine, and re-using the brine. Desirably, the brine comprises cesium as the majority alkali metal by weight, and formate, acetate or other species, as the salt anion. The brine may comprise a significant proportion of other alkali metals, and may therefore contain mixtures of alkali metal carboxylates. Further, the brine may comprise one or more polymers or other components which are adjuvants and provide desirable properties to the brine or avoid disadvantages.

The preferred silver salt is silver nitrate, and for ease of description, all references hereinafter will be to silver nitrate.

Since the specific gravity of the brine is extremely important, it is desirable to minimise any loss of specific gravity by undue dilution during the treatment. Accordingly, it is desirable to use silver nitrate solutions containing at least 200g/l of AgNO3, more preferably at least 300g/l AgNO3 and most preferably at least about 800g/1. At room temperature, saturation concentrations are about 1400g/l AgNO3. The silver nitrate solution may contain other components which do not significantly adversely interfere with the method of the invention, or the performance of the brine. The silver nitrate solution is conveniently a product stream from a process involving the manufacture of high purity silver nitrate. This can provide economies arising from heat and water savings and other processing costs.

The treatment of the invention is conveniently carried out at room temperature, but may be carried out at higher or lower temperatures. It will be borne in mind that solubility decreases with decreasing temperature, and crystallisation may occur. Depending upon the

chloride concentration, it may be necessary to cool the brine to about 0°C, to facilitate the removal of sufficient nitrate salts formed as a by-product to create a stable solution that can be supplied and used in winter conditions.

Conveniently, silver chloride may be removed from the brine by filtration. This is particularly applicable because of the relatively high value of the silver chloride. Other methods for separation may be used, however, such as hydrocyclones or centrifuges, and where applicable or desirable, polymeric materials may be added to enhance separation.

The treatment may be carried out in a two-step process, or, in a preferred embodiment, in a single step process. Desirably, at least a 90% stoichiometric quantity of silver nitrate is used, relative to chloride ion, in the treatment. More desirably, the quantity of silver nitrate is from approximately 95% to approximately 112% stoichiometric, for brines containing 8.47 to 13. 5g/l chloride. If the brine has other chloride contents or components, the stoichiometric adjustment may differ, which can be established by trial and error.

The recovered silver chloride may carry entrained by-products such as silver formate and cesium nitrate. Under preferred conditions, these are minimised and removed to permit the production of a re-usable brine. By washing and crystallisation, silver chloride crystals may be obtained and the silver value recovered in conventional ways by conversion to other compounds or silver metal, using methods available to the person of ordinary skill in the art.

In general, therefore, it is preferred to separate the silver chloride by filtration, but conventional washing of the solids is not desirable if it causes undue dilution of the filtrate.

Entrained cesium formate in the filtered precipitate may be washed out subsequently, using water, and cesium nitrate may also be recovered if desired.

The skilled person may use the information herein to optimise the process, using conventional techniques.

The invention may be further understood with reference to the following Examples.

EXAMPLE 1 Stoichiometric Additions of AgNOß 200 ml samples of a used cesium formate brine, containing 1587g/l cesium formate, 13. 53g/l chloride, 0.720wt% sodium and 2.63 wt% potassium were used for all tests described. Desirably, the chloride level will be reduced below lg/1, more preferably 0.3-0. 7g/l.

In the first Example, stoichiometric amounts of AgN03 solution are added with stirring, to the brine, at room temperature and at differing concentrations: AgN03 concn. (g/1) 150 200 600 918 1366 Volume added (ml) 87.6 65.7 21.6 14.5 9.5 Cl-concn. (g/l) 0.50 0.64 0.72-0. 99 1.0 1.78 (all concentrations in g/1 normalised to 200 ml) Specific gravity 1.848 1.913 2.084 2.117 2.129 In order to produce a product brine having Cl-concentration of approximately 0. 5g/l, combined with a specific gravity of not less than 2.0, a second stage treatment with AgN03 was undertaken, with the following results: First stage product Cl- (g/1 normalised) 0.99 1.00 1.78 AgN03 stoichiometry (%) 50 50 80 AgN03 concn. (g/1 normalised) 200 90 200 Final Cl-concn. (g/1 normalised) 0.54 0.51 0.44 Overall AgN03 stoichiometry (%) 103. 2 103. 4'110. 0 Specific gravity 2.055 2.077 2.057 After first and second stages, the deposits formed in the brine were filtered off.

EXAMPLE 2 Single Stage Chloride Removal The identical cesium formate brine as in Example 1 was used in further tests, using differing stoichiometries: Test No. 1 2 3 AgN03 stoichiometry 107 110 116 Vol. AgN03 soln. added (ml) 10.2 10.4 11 Cl-concn. (g/l normalised) 0.58 0.35 <0. 01 Specific gravity 2.111 2.122 2.095 Tests 1 and 2 proceeded satisfactorily at room temperature. It was assessed in Test 3 that excess silver was being dissolved and a post-treatment of heating the product brine to 95°C was incorporated, to remove the silver in solution.

A further post treatment of cooling to approximately 0°C overnight followed by filtration, was found to remove a large proportion of by-product cesium nitrate, leaving a stable clear solution at room temperature.

EXAMPLE 3 Single Stage Chloride Removal from a K/Cs Formate Brine 200ml samples of a potassium/cesium formate brine, containing 497g/l formate and 8. 47g/l chloride, 0. 675wt% sodium, 12.56wt% potassium and 33.9wt% cesium were used for all the tests described in this Example 3. Desirably, the chloride level will be reduced such that it lies in the range 0.7 to l. Og/1.

An orthogonal array of sixteen tests in which different concentrations and stoichiometric amounts of AgNO3 are added at different temperatures, with different stirring

rates, addition rates and residence times is shown below. In the tests in this Example, the reactants were stirred with an IKA Werke RCT Basic stirrer, using either a"Slow"setting (setting 4) or a"Fast"setting (setting 7). Addition rate was either"Slow"or"Fast", corresponding to a 15 minute addition time or a 30 second addition time, respectively.

Residence times are the times from addition of the last drop of AgNO3 to the beginning of filtration, and"Short"means 5 minutes and"Long"means 30 minutes. AgNO3 Stoichio-Temp-Stir Addition Residence Norm. Specific conc. metry erature Rate Rate Time Cl-conc. Gravity (g/l) (%) (°C) (g/1) 600 95 25 Fast Fast Short 0. 68 1. 859 600 100 25 Slow Fast Long 0. 81 1. 867 600 105 35 Fast Slow Long 0. 01 1. 847 600 110 35 Slow Slow Short 0. 38 1. 848 800 95 35 Slow Fast Long 1. 61 1. 867 800 100 35 Fast Fast Short 0. 42 1. 871 800 105 25 Slow Slow Short 0. 71 1. 856 800 110 25 Fast Slow Long 0. 01 1. 851 1000 95 25 Slow Slow Long 1. 86 1. 856 1000 100 25 Fast Slow Short 0. 41 1. 857 1000 105 35 Slow Fast Short 1. 37 1. 871 1000 110 35 Fast Fast Long 0. 01 1. 868 1200 95 35 Fast Slow Short 0. 87 1. 884 1200 100 35 Slow Slow Long 1. 84 1. 886 1200 105 25 Fast Fast Long 0. 14 1. 871 1200 110 25 Slow Fast Short 3. 26 1. 865

By following the orthogonal matrix, it can be demonstrated that the chloride level can be reduced from 8. 47g/l to between 3.26 and <0. 01 g/1 by the addition of AgN03 under various conditions.

Analysis of the chloride levels observed by following the orthogonal matrix suggests that the stirring rate, concentration and stoichiometry of the AgN03 added, are the most significant factors that determine the post-treatment chloride level in descending order of importance. In particular, a fast stirring rate is highly desirable for efficient AgCl precipitation as the formate reduction of silver side reaction is minimised. The temperature, addition rate and residence time appear to be less significant factors.

To validate the conclusions drawn from the orthogonal array, a further three confirmation tests were designed specifically to reduce the post-treatment chloride level to 0.85, 0.65 and 0. 45g/l. The same brine was treated at 35°C, with the AgNO3 added slowly with fast stirring before a long residence time.

Test No. 1 2 3 AgNO3 stoichiometry (%) 95 95 100 AgN03 concentration (g/1) 1200 800 1000 Target Cl-level 0.85 0.65 0.45 Cl-concentration 0.88 0.66 0.37 (normalised, g/1) SG 1.890 1.874 1.884 The most AgNO3 efficient reduction in Cl-level from 8.47 to <lg/l and the lowest amount of water added occurs in Test 1.