WO1998039257A1

WO1998039257A1 - Method and equipment for the evaluation of scale dissolvers

Info

Publication number: WO1998039257A1
Application number: PCT/NO1998/000055
Authority: WO
Inventors: Kari Ramstad; Trine Tydal; Norvald Gjelsvik
Original assignee: Norsk Hydro Asa
Priority date: 1997-03-03
Filing date: 1998-02-25
Publication date: 1998-09-11
Also published as: NO970958D0; AU6312498A

Abstract

Method for the evaluation of scale dissolvers using factorial design static tests. The scales subjected to dissolution by the dissolvers are manufactured as pellets on the basis of scale powders in the desired mole ratio. More particularly the scale powders are mixed with distilled water to form a paste which is subjected to pressing in a pellet press, which formed pellets are thereafter dried and sintered. An apparatus for performing dynamic flooding tests on the basis of scale pellets, for the evaluation of scale dissolvers, includes a flow cell comprising a basically cylindrical scale pack (1) with end pieces (2), each including intermediate filters (3), which end pieces (2) are provided with throughgoing bores and fittings (5) for the connection of supply and outlet pipes (6).

Description

"Method and equipment for the evaluation of scale dissolvers"

Mineral scaling is a well known phenomenon in the oil and gas production industry. Carbonate scale may form during production of connate water or formation water, while sulphate scale may be formed by mixing of incompatible waters, such as formation water and injected seawater. If these waters mix downhole, barium sulphate and strontium sulphate deposits may cause plugging in the near wellbore area and perforations leading to production loss. These salts often occur as the co-precipitate (Ba, Sr) S0₄ often having traces of RA in the lattice, which makes the scale radioactive. Calcium sulphate scaling and formation damage due to loss of Ca-based brine followed by seawater breakthrough has been reported. If formation water and seawater are produced from different wells and mix at the manifold, the topside facilities can suffer from these hard sulphate scales.

Commercial scale dissolvers may be used in the near wellbore area, in the wellbore or in the topside equipment to dissolve or soften the scale deposits and to make removal more easy. Carbonates are acid soluble and inorganic or organic acids are regularly used for this purpose. Sulphate scales, on the other hand, are very hard to dissolve and require special chemicals, such as sequestering agents. The conditions under which these chemicals are used will determine how successful a treatment will be. The optimum conditions may vary depending on type of scale. The objective of the present invention has been to establish laboratory methods for evaluation of scale dissolvers and to find the significant factors to optimise a scale dissolver treatment.

Another objective has been to develop apparatus for the evaluation of scale dissolvers.

The invention is defined in the attached independent method claim 1 with dependent claims 2-4, and independent apparatus claim 5

EVALUATION OF SCALE DISSOLVERS

General

Dissolution of sulphate scale requires chemicals such as sodium carbonate/bicarbonate, ammonium carbonate/bicarbonate and salts of ethylene diamine tetraacetic acid (EDTA), diethylene triamine pentaacetic acid (DTPA), nitrilotriacetic acid (NTA), hydroxy acetic acid and gluconic acid. These chemicals or mixtures of them work as converters or sequestering agents (chelants) and are as a common name referred to as scale dissolvers. The use of synergists together with the traditional sequestering agents have been proved to enchance the kinetics of the reaction. Potassium salts of the chelants have been found to be especially effective towards barium sulphate. Some of the dissolvers convert the scale to a different solid that is more easily removed or dissolved.

Several workers have reported studies of the dissolution of sulphate scales in static or semi-static tests. These experiments are often based on the dissolution of drilling mud grade barite or small scale chips in scale dissolvers. This is a reproducable test, but does not represent the real case, where the ratio of the scale surface area to the volume of the scale dissolver is much less and the dissolution rate is much slower. Real scale samples which may provide the correct surface area to volume ratio, will on the other hand vary in composition and can not be used as a standard medium.

These limitations called for a new kind of scale samples that could overcome this problem and be used in static experiments for evaluation of scale dissolvers. Further, the effect of flooding scale dissolvers through a porous medium containing scale has not yet been fully investigated, although reported once.

Well treatment to be simulated in static tests

A treatment of a wellbore tubing (200 m, ID = 0.1597 m) covered with a sulphate scale layer of 1-2 mm thickness was to be simulated in static tests. Typical North Sea scale is BaS0₄:SrS0₄ 6:1 with a typical density of 3.5 g/cm³. This gives a surface area to volume ratio of 1 cm² : 4 cm³.

In order to cover different types of scale in the study, it was decided to perform static tests on mixtures of CaS0₄ and BaS0₄. SrS0₄ was assumed to have characteristics somewhere in between the two scales and the results could be interpreted for this type of scale too, see below.

The scale dissolver solution is usually bullheaded into the wellbore. The well is shut in to allow soaking and good contact between the dissolver and the scale. During the shut in the solution is pumped further in steps to make sure that fresh dissolver is in contact with the scale. This last situation is difficult to simulate in a static test, but was studied separately in dynamic flooding through a porous medium, see below. Agitation will have some effect on the dissolution process. The temperature of the scale dissolver solution will normally increase during a well treatment. The injection fluid may have a low temperature (normal temperature of injection water is 4-28 °C) and the near wellbore area temperature may be over 100 °C. During the shut in the dissolution rate may therefore increase. In order to shorten the necessary treatment time, the dissolver may be injected as a preheated solution. We wanted to study this behaviour and a temperature range of 40-80 °C was examined.

Shut in time is a recurring issue when designing dissolver treatments. The operator of a field wants to minimize the shut down time and do a treatment as quickly as possible. The chemical, on the other hand will need some contact time with the scale to work in an optimum way. This is a challenge for the vendors and several commercial products are now containing synergists which are claimed to shorten the time needed for dissolving a certain amount of scale. Service companies recommend shut ins lasting from 12 hrs to several days, depending on the dilution of the dissolver and the opportunity to do the treatment in steps by adding further fresh dissolver. We wanted to look at the effect of shut in time between 4 and 20 hrs.

Dilution of the scale dissolver may occur during the treatment or may be chosen initially to cut costs. The effect of dilution upon the dissolver efficiency was therefore studied. The dissolvers were examined as concentrated solutions (100 %) and 50 % diluted in 2 % KCI, a medium which is reported by several workers. Some service companies often recommend even more diluted solutions than 50 %.

During back production after a scale dissolver treatment in a well, the pH of the dissolver solution will decrease from alkaline to the neutral or low pH of the produced water. We wanted to simulate this change and observe the effect of the scale dissolvers as a function of pH.

The dissolution of scale in the near wellbore area may be governed by different factors than the static soaking situation. Narrow pore throats may lead to less favourable flow patterns in the porous rock. This was studied in a dynamic sandpack flooding experiment.

Experimental design

Based on the parameters listed above static tests were performed to simulate the well treatment. This was done by using a 2-level statistical factorial design to investigate which factors and factor interactions that are important, and to which extent they influence on the scale dissolution process.

The factors of the design were:

- Ca content in the scale (as mole % of the total amount of Ca and Ba moles)

- temperature in the solution during soaking

- soaking time

- concentration of the scale dissolver solution

- initial pH in the solution

The high and low levels of the factors and the measured responses are presented in Table 1.

The high value of the pH in the dissolver solution was chosen as that of the supplied dissolver (100 %) or that given after dilution in KCI (50 %), while the low pH was 7.5 adjusted by addition of cone. HCI. The center point was the mean value of the high and low level and would therefore vary from dissolver to dissolver depending on the high pH. Some dissolvers already having a neutral pH were not further acidified, but examined only at one value of pH, namely as supplied.

Three of the responses from the experiments were quantitative measurements. These were the total loss of weight of the scale (%), the amount of Ca ions dissolved into the solution (%) and the amount of Ba ions dissolved into the solution (%). Some of the scale dissolvers were containing Ca in the solution (up to 360 mg/l). This had to be corrected for in the calculation of the residual Ca content. The Ba level in the dissolvers was very low and was not taken into account. The combination of the loss of weight and ions dissolved will give some information on the dissolver's ability to convert the scale.

The other two responses were qualitative. One was defined as the amount of particles on the filter after filtration of the dissolver/scale sample (Filt 0-3), see Experimental below. 0 means no particles in the filter, through to 3 which means many particles. The second qualitative response was defined as the condition of the scale pellet after the treatment (Tabl 0-3), where 0 means that the surface of the scale was unchanged, 1 means a rough surface, 2 means that the scale had become porous and 3 means divided specimen. These visual observations gave a new kind of information about the scale dissolvers that is not observed when studying the dissolution of powder scale. This makes our method very useful when considering whether mechanical removal of the scale (milling or water jetting) will be easy to perform after soaking of the scale.

Table 1 Factors and responses in the experimental design of the static tests

Table 2 2.5-1 experimental design with 3 centre points. + high level, - is low level, 0 is center point.

The experiments were performed using a 2^5"1 design with 3 center points, see Table 2. The design includes 16 experiments per dissolver varying the factors at low and high levels and 3 experiments using the mean values of the low and high levels. The center points are replicates and will determine the standard deviation of the set. The results from the experimental design have been used to create models for each scale dissolver.

EXPERIMENTAL

Static tests of artificial CaSO₄/BaSO₄ scale

Gypsum and barite (both drilling mud quality) were weighed to give the desired mole ratio and mixed as powders. A small amount of distilled water was added and the paste was pressed to form pellets (i.e. under 1 ton/sqi for 1 min, 5 tons/sqi for 1 min and 10 tons/sqi for 1 min). Different amounts of solid were tried and the most consolidated pellets had a diameter of 13 mm and a hight of 3 mm. Figure 1 shows a photograph of an artificial CaS0₄/BaSO₄ scale pellet. The tablets were dried in a desiccator 16 hrs (room temperature), in an air bath for 1 hr (150 °C) and carefully sintered for 16 hrs (1000 °C). This gave dense pellets with a surface area of about 4 cm² and a porosity of 24 %, which fairly represents a real scale sample. In the static tests the surface area to volume ratio was obtained using a dissolver volume of 16 ml.

The weight of the pellets were 1-1.5 g. Figure 2 shows a scanning electron micrograph of an artificial CaSO₄/BaS0₄ scale pellet. The micrograph reveals a conglomerate containing local gains of CaS0₄ and BaS0₄ in the mixed matrix (enlarged 270x). Element analyses showed traces of Al and Mg. The heating and sintering removed all the pore water and crystal water.

The stoichiometric ratio of dissolver to Ca+Ba was 3.2:1 for the 100 % dissolver solution and 1.6:1 for the 50 % solution based on EDTA.

16 commercial scale dissolvers were chosen for the study. These were mostly polyaminocarboxylic acids, some of them containing synergists. Almost every one of the scale dissolvers had a pH above 11. Some of the products were designed for CaS0₄ scale and had a pH around 7. For comparison, a standard 40 % EDTA solution (pH=12.8) was examined.

Static tests were performed according to the experimental design, but in random order, to avoid any systematic errors. The scale dissolver solution was prepared in a 25 ml glass bottle with a screw cap and preheated over night in a hot cabinet. The dry pellet was weighed and added to the bottle which was closed and left in the hot cabinet for the desired soaking time. The bottle was tilted every hour in the 4 hrs experiments and about every second hour in the 20 hrs experiments to simulate the motion available in a field treatment procedure. The test was therefore semi-static.

After the treatment, the warm solution was filtered (0.45 μm) at constant temperature using a water bath. The bottle was then washed with the filtrate and the solution was filtered again.

The solution was diluted with distilled water to provide a proper sample for ICP (Inductive Coupled Plasma) element analyses of Ca and Ba. The filter was washed with distilled water (2 ml) and dried for 16 hrs at 80 °C. The pellet (or what was left of it) was dried separately for 16 hrs at 190 °C. This was done to remove any pore water or crystal water regenerated during soaking. The pellet was then weighed to determine the loss of weight.

Static tests of real (Ba,Sr)S0₄ scale

(Ba,Sr)S0₄ real scale was supplied from Field 1 in the North Sea, sample taken at the inlet of the interstage cooler in the topside equipment on the platform. In some cases real scale samples have been rinsed in xylene prior to a laboratory testing (3). This is not always possible to do in the field and we chose not to treat the scale sample any further. The sample was characterised by X-Ray Diffraction and Scanning Electron Microscope (XRD/SEM). When Ca (1 %) was ignored, the composition of the scale was Ba₀₇₄Sr₀₂₆SO₄.

Another sample was provided from Field II in the North Sea. The sample was removed from pall rings by water jetting. This contained traces of hydrocarbons. No pre-treatment of the sample was undertaken. The scale was characterised by XRD and SEM analyses. XRD analysis showed a solid solution of Ba and Sr sulphate in the same lattice. The composition of the scale was determined to Ba₀₅₃Sr₀₄₇SO₄, if other metals were ignored (Ca 2 %).

Static tests in a 2^4"1 factorial design were performed on the real scale following the same procedure as for the artificial scale, omitting the variation in the scale composition (CaBa) in the model. Including 3 center points this led to 11 experiments per scale type per scale dissolver. Five chemicals containing mainly polyaminocarboxylic acids were examined. Details about the composition of the dissolvers were not available. The dissolvers were abbreviated as A, D, E, F and S.

Dynamic tests of synthetic (Ba,Sr)S0₄ scale

Dynamic flooding tests of the five dissolvers A, D, E, F and S were performed using a specially designed cell as is shown in Figure 3. The cell comprises an acrylic plastic cylinder 1 for the pellets 7 to be tested, end pieces 2 of titanium with sealing rings 4, filters 3, inlet and outlet teflon tubes 6 being connected to the end pieces 2 by means of stainless steal fittings 5, and a clamp arrangement to hold the cell together. A space between the filters 3 and pieces 2 provides even distribution of the flooding medium along the cross section at the inlet as well as at the outlet of the cell.

The concentrations of the dissolvers, were 100, 100, 25, 50 and 50 % respectively in 2 % KCI. These concentrations were in agreement with the suppliers recommendations, except for E where fresh water was suggested as a diluent Silica sand with a mean particle size of 160 μm was chosen as the inert host matrix for the dynamic scale dissolver experiments

In situ scale development in a sandpack may be difficult to reproduce (Ba,Sr)S0₄ synthetic scale was therefore made in the form of small particles by precipitation This was carried out by carefully mixing aqueous solutions of BaCI₂ 2H₂0 p a and SrCI₂ 6H₂0 p a and Na₂S0₄ at 80 °C with a stoichiometric ratio of Ba:Sr:S0₄ of 1 :1 :2 Drilling mud quality was not suitable for this part of the study, due to time consuming preparation The XRD diffractogram of the scale showed that a solid solution of (Ba,Sr)S0₄ had been formed by the precipitation The particle distribution was measured and showed a mean value of 2 4 μm

The flow cell was filled with a wet paste of the synthetic scale (15 wt%) and sand (85 wt%) using 2 % KCI as the wetting agent This gave the most even mixing and density The packing procedure gave reasonable reproducabi ty The pore volume was 22 ml To avoid later collapse of the sand/scale pack, the cell was tapped slightly towards a bench plate during the packing, and excess of KCI was allowed to drain out through the bottom end piece Finally the cell was clamped together with two parallel plates joined by four screwing bolts

2% KCI was flooded (upwards) through the cell for 30 mm to fill possible voids in the particle bed (3 ml/mm) The cell was heated to 80 °C during KCI flooding at 0 5 ml/mm for 85 mm and at 3 ml/mm for 15 m The cell was then flooded with a flow rate of 3 ml/mm under continuous collection of ca 15 ml fractions by the following procedure 150 ml of scale dissolver, four cycles with 30 mm shut in, and 15 ml scale dissolver flooded (no shut in after the last 15 ml batch), 150 ml KCI flood (same direction) The cell was then disassembled and the sand/scale plug was carefully pushed out for visual inspection. The effluent fractions were diluted 1 :1 with EDTA-solution and analysed for Ba and Sr by ICP. Two replicates were run for each dissolver.

For comparison, saturation tests of crushed synthetic (Ba,Sr)S0₄ scale were performed with the same five scale dissolvers at 80 °C, 48 hrs, no agitation, to determine the maximum dissolution capacity of the dissolvers.

RESULTS AND DISCUSSION

Modelling based on CaSO₄/BaSO₄ scale

As an example, the results from the CaS0₄/BaS0₄ pellet static tests for scale dissolver S are shown in Table 3. The dissolver shows a high degree of dissolution efficiency (weight and ions). Further, the qualitative responses show that dissolver S has a high ability to soften and divide the scale which will make it more easy to remove mechanically after soaking. The center points of the design show a convincing reproducability in the experimental procedure.

Modelling of the 2^5"1 design was performed using MODDE 3.0 from Umetri, Sweden. The statistical results for scale dissolver S are shown in Figure 4. As can be seen from "Summary of Fit", the model is very good in describing dissolution based on loss of weight, Diss, (fraction of the variation of the response explained by the model, R2, is equal to 1 ). R2 is also high for the other responses, indicating an overall good model. The ability to predict from this model is best for Diss and Tabl (dividing of pellet), however, ion analyses and Filt (particles on the filter) may give less accurate prediction (predictive power of the model, Q2, is much less than 1 ). This could be due to outliers or some lack of linearity in the data set and might be improved by refining of the model. The model for weight reduction (Diss) of this particular scale dissolver can be used as an example of all the dissolvers, although it will not represent the characteristics of the other chemicals. The model including significant factors (i.e. coefficients greater than 3 std.dev.) is as follows:

% Diss = 15.6 + 10.3 Ca + 4.6 T - 4.2 C - 3.2 Ca*C

The model shows that a high Ca content in the scale will promote the dissolution process. Furthermore enchanced temperature will be favourable and the dissolver will work best as a diluted solution (negative contribution from concentration). There is a mixed effect of Ca and C showing that these parameters ought to be considered together. As an example the contour plot of Diss as a function of CaBa and Temp is shown in Figure 5. As can be seen, high temperature has a significantly positive effect on dissolution of scale with high Ca content, but is less important for scale with low Ca content.

Similar models on the three quantitative responses were developed for all the dissolvers. The scale dissolvers showed quite different characteristics. The overall conclusion was that the presence of Ca in the scale occupies a major capacity of the scale dissolver. The dissolver will complex with every metal that is present in the order of stability. This means that Ca-scale e.g. calcium carbonate ought to be removed (acid) prior to appliance of the scale dissolver. The well or plant must, however, be flowed to neutral condition before scale dissolver is introduced if BaS0₄ is to be dissolved. The amount of Ba ions released into the solution is dependent on pH. All the alkaline products caused higher release of Ba ions into the solutions at high pH. This could indicate that during back production of water after a well treatment, the bulk pH may decrease and lead to less dissolution of BaS0₄. Table 3 Results of the CaS0₄/BaS0₄ pellet static tests for scale dissolver S.

12 of 16 commercial dissolvers approved upon dilution, most of them on the total amount of weight reduction and some of them on the amount of Ca ion dissolution. In most cases this is in agreement with the recommendation from the service companies. However, this demonstrates the necessity of doing laboratory studies to optimise the treatment conditions prior to field appliance.

14 of the dissolvers including the EDTA standard benefited from long soaking time, especially for Ca dissolution. 11 of them also had a good effect of high temperature. This means that preheating of the solution may improve the result of the dissolver job.

The scale dissolvers may be ranged based on the average amount of scale dissolved (weight) in all the experiments. The 10 overall best products in this study were able to reduce the weight of the scale by 12.5-16.5 % as an average of all the static tests. The maximum reduction of weight was over 50 % for a pellet with high Ca content. These dissolvers were polycarboxylic acids with or without synengist and more exotic additives. Six products gave an average weight reduction of 4.3-9.7 %. One of the dissolvers was a typical converter, which gave very low reduction in weight, but the Ca ion release was rather high.

The highest average dissolution of Ca ions was 58 %, obtained by one of the products designed for gypsum scale. This chemical gave a very poor Ba dissolution.

The majority of the dissolver models included significant 2-factor interaction effects from the combinations of the Ca, pH, Ti and T. This shows that one of the factors in combination with another can have a significant effect on the result even if the factor itself is insignificant.

Prediction of real scale dissolution using CaS0₄/BaS0₄ models

The results from the static test on the real scale samples from Field I and II (2⁺¹ design) were compared to the predicted dissolution using the 2^5"1 statistical models. The five chosen scale dissolvers were A. D, E, F and S.

Since the models had been created based on CaS0₄/BaS0₄ synthetic scale, the prediction of dissolution of (Ba,Sr)S0₄ using these models had to be performed with a weighed Ca/Ba ratio as an input. The solubility product (K_. ) of the sulphates are for CaS0₄ 2.4 ^• 10^"5, SrS0₄ 7.6 ^• 10^'7 and BaS0₄ 1.5 ^• 10^_9(7). Further, the stability constants (log K at ion strength 0.1 , 20 °C) for the corresponding metal-EDTA complexes are Ca-EDTA 10.7, Sr-EDTA 8.6 and Ba-EDTA 7.8(8). This means that the solubility of the Sr-salts can be predicted somewhere in between the Ca and Ba values.

Based on the XRD/SEM analyses of the real scales from Field I and II, weighed CaBa mole % distributions were chosen as given below (CaBa of 12 means 88 % Ba in the prediction):

Ba Sr CaBa prediction values

Field I 74 26 12

Field II 53 47 25

The measured and predicted results on weight reduction (Diss) and Ba ions in solution (Ba) of the different experiments of scale dissolvers A and S are plotted in Figures 6 and 7, respectively.

The predictions of dissolved scale from Field I in scale dissolver A are in very good agreement with the measured values.

The scale from Field II, on the other hand, seems to be easier to dissolve than predicted in dissolver A, but the trends in the various experiments are the same. This is reflected both in the weight dissolved and in Ba ions in solution. This could mean that the scale is very porous or that the CaBa prediction value is weighed too close to BaS0₄. However, the CaBa prediction value had to be raised to 50 % Ca, to give a predicted dissolution of 31 % (measured was 35 %) and a Ba ion dissolution of 17 % (measured 37 %) in Exp. no. 7. In Exp.no. 11 the corresponding value gave 14 % dissolution (measured 15 %) and 12 % Ba released (measured 35 %). The model is conservative towards predicting dissolution of the Field II type scale with a relatively high Sr content. Further, we see that almost the same amount of Field II scale is dissolved at 40 °C, 20 hrs soaking, as is dissolved at 80 °C, 4 hrs soaking. (Exps. no 11 and 13, 100 % solution, high pH). This means that preheating of the solution may shorten the time needed for cleaning of equipment from sulphate scale deposition.

For scale dissolver S there is quite good agreement between the predicted and measured dissolution of the Field I scale. For the Field II scale predicted values are lower than the experimental data. Some divergence in the center points could reflect the heterogenity that may occur in real scale.

Predictions of Field I scale in dissolver D were also very close to the measured values, while dissolver E gave less consistent results (not shown here). For scale dissolver F dissolution is slightly overestimated by the models. This is seen very clearly in the Ba results.

The models for dissolvers D, E and F were all conservative towards the scale from Field II (scale was again more easily dissolved than predicted).

Overall, the scale dissolvers can be ranked based on the loss of weight (Diss) of the Field I and II scales at the most favourable conditions, namely 80 °C, 20 hrs soaking, 50 % solution and high pH (Exp no. 7). The order is valid for both Field I and Field II type of scales ('>' is better than ):

D > F = S > A > E

The overall conclusion is that the CaS0₄/BaS0₄ models can be used to predict (Ba,Sr)S0₄ dissolution at given conditions. The models are somewhat conservative for the sulphate scale type with relatively high Sr content, but may be improved by adjusting the CaBa input value. In the real case offshore, the models may be a useful tool for predicting the effect of scale dissolvers, especially topside where the conditions can be controlled in a much better way than in the downhole treatments.

Effect of dynamic flooding on synthetic (Ba,Sr) S0₄ scale

The dissolution profiles (Ba and Sr release) of the dynamic flooding tests of the five dissolvers A, D, E, F and S are shown in Fig. 8. The fact that the sandpack contained a known amount of (Ba,Sr)S0₄ scale made determination of % dissolution of these ions possible. For each dissolver, two parallel runs were performed. The 30 min shut ins of the scale dissolvers are marked with '+'.

The effect of shut in seems to be small and contributes probably less than 5 % to the total amount dissolved (estimated from the profile and the slope of the curve). This is probably due to formation of channels prior to the shut in. This is indicated by the profiles showing a Ba and Sr concentration peak during the initial dissolver flooding step. The position of the peak is in the region 50-90 ml for the most efficient dissolvers, i.e. when only 10-15 % of the barium is dissolved. Formation of channels was confirmed by the visual inspection of the plug after the flooding, see below. Therefore, only marginal further dissolution is gained by shutting the dissolver in, as most of the scale originally present where the channels formed was already dissolved.

The above shows that dissolver flooding is important. As channels form, constantly increased dissolver flooding rate is probably necessary to obtain the pressure needed to force the dissolver into new areas. During shut in, no new areas will be reached by the dissolver. The dissolver will probably go out through the channels and out of the cell largely undersaturated with dissolved scale. This demonstrates the importance of movement and possibly supply of fresh dissolver for treatments in porous media. The inspection of the sand/scale plug after the dissolver/KCI treatment had been completed showed two distinctly different areas. One was a dark shaded area having a cylindrical shape and found mainly in the lower part and outer radius regions of the cell where the scale had been dissolved. The other was a light shaded area corresponding to sand/scale (compositions of the two different areas were analysed). This shows that heterogeneous dissolution takes place in the porous medium. The patterns indicate that the dissolver 'digs out' channels in the porous structure, due to the increased permeability resulting from the dissolution process.

The preference of the dissolvers towards dissolution of barium and strontium in the dynamic tests are also shown in Figure 8. For all the dissolvers, more strontium than barium dissolves, and strontium dissolves more rapidly (steeper initial dissolution curve). However, the ratio of dissolved Ba and Sr varies considerably. As seen in the plots, more Sr than Ba also dissolves during the finishing KCI flooding, showing that Sr dissolves easier in diluted dissolver.

There was good agreement between the results from the dynamic tests and the corresponding saturation tests on crushed synthetic scale. In the dynamic tests, however, the ratio of Ba/Sr dissolved is higher than in the static tests. In the static tests Sr dissolve more rapidly than Ba, hence occupying the major capacity of the dissolver. Under dynamic conditions, the constantly supplied fresh dissolver will dissolve BaS0₄ initially left behind when Sr is washed out (steep curve for Sr). Due to the formation of channels, less Sr is available for the dissolver.

Dissolver D was found to be particularly efficient in the dynamic tests. At the given conditions, the five dissolvers may be ranked from the dynamic tests as follows ('>' is better than):

Ba dissolution: Sr dissolution:

D > A > F > S > E D > E > A > F > S The static tests can be performed within short time and at minimal cost, and is therefore specially suited to study the effect of different treatment conditions. The dynamic tests, however, have revealed additional information concerning flooding and shut in, which is important to take into consideration, especially for design of field operations.

CONCLUSIONS

During the present laboratory study of scale dissolvers the following conclusions have been made:

i) The effects of scale composition, temperature, treatment time, dissolver concentration and pH on the scale dissolution process have been determined for a range of commercial dissolvers using factorial design static tests. Such information is crucial for efficient scale dissolver treatments.

ii) Static tests using artificial sulphate scale pellets provide a more realistic ratio of scale surface area to dissolver volume. The visual observations of the pellets make the method useful when considering whether the scale can easily be removed mechanically after soaking.

iii) The composition of the scale (Ca/Ba/Sr ratio) will be of great importance to the scale dissolver performance. Mineralogical analyses should therefore be performed prior to dissolver selection.

iv) By using experimental factorial design in laboratory studies significant higher order factor interactions can be revealed.

v) The statistical models developed can be useful tool for predicting the effect of a scale dissolver at a given set of parameters. The models are somewhat conservative for the sulphate scale type with relatively high Sr content, but the prediction may be improved by adjusting the CaBa input value.

vi) The dynamic technique developed gives reproducable results and is suitable for studies of dissolver flooding and shut in an initially homogeneous sand/scale environment.

vii) During dissolver flooding heterogenous dissolution occurs, forming channels in the porous medium. A constantly increased dissolver flooding rate is probably necessary to obtain the pressure needed to force the dissolver into new scaled areas.

Claims

1. Method for the evaluation of scale dissolvers using factorial design static tests, characterised in that the scales subjected to dissolution by the dissolvers are manufactured as pellets on the basis of scale powders in the desired mole ratio.

2. Method according to claim 1 , characterised in that the scale powders are mixed with distilled water to form a paste which is subjected to pressing in a pellet press, which formed pellets are thereafter dried and sintered.

3. Method according to claim 2, characterised in that the pellets are dried in air at approximately 150┬░C and thereafter sintered at approximately 1000┬░C.

4. Method according to claims 1-3, characterised in that the pellets are manufactured with a size having a diameter of approximately 13 mm and a height of approximately 3 mm.

5. Apparatus for performing dynamic flooding tests on the basis of scale pellets, for the evaluation of scale dissolvers, characterised in a flow cell comprising a basically cylindrical scale pack (1) with end pieces (2), each including intermediate filters (3), which end pieces (2) are provided with thoroughgoing bores and fittings (5) for the connection of supply and outlet pipes (6).