Impact of In-Situ Sulfate Stripping on Scale Management in the Gyda Field

Abstract

While barium stripping is commonly observed in sandstone reservoirs where seawater mixes with formation water that may be rich in calcium, strontium and barium ions, this paper presents evidence for in situ sulphate stripping in a sandstone reservoir. The formation brine composition suggests that a moderate to severe barite scaling tendency will require inhibitor concentrations in the range of 10–50 ppm to control scale, but in practice concentrations < 5 ppm are adequate. Investigation of the produced brine compositions has revealed that this is due to much lower sulphate concentrations in the produced brine mix than would be expected purely from dilution of seawater with the formation brine. The question this paper addresses is what has caused this reduction in sulphate concentration. The formation brine Mg/Ca ratio is < 0.1. Over geological time frames, the reservoir rock and formation brine will come into chemical equilibrium, the Mg/Ca and Na/Ca ratios in the brine being dependent on the respective ratios in the rock matrix. However, when seawater is injected, this equilibrium is disturbed. Since the Mg/Ca ratio for seawater is ∼ 3, to re-equilibrate an ion exchange mechanism causes magnesium to be retained from the brine phase onto the rock, and in return calcium is released from the rock into the brine phase. This is confirmed by lower than expected magnesium concentrations in the produced brine. The impact of the calcium release into seawater as it is displaced through the hot reservoir is to cause precipitation of calcium sulphate, this process resulting in the observed sulphate stripping. This analysis is supported by the field data and by reactive transport calculations. Implications are drawn for scale management in this and similar fields with high formation water calcium concentrations. Introduction The Gyda field lies on the north-eastern margin of the North Sea Central Trough, on the Norwegian Continental Shelf, 270 km (168 miles) southwest of Stavanger and 43 km (27 miles) northeast of Ekofisk Centre. The offshore installation comprises a conventional 6-legged steel jacket which supports integrated production, drilling and living quarters. Peak oil production topped 20,100 m3/day (126,000 stb/day) during 1993. Gyda is currently operated by Talisman-Energy Norge A/S (61 %) on behalf of DONG (34 %) and Norske AEDC A/S (5 %). It was originally operated by BP Norway Ltd., and when it came on stream in July 1990, it was the deepest, hottest and lowest permeability oilfield in the North Sea[1]. Gyda receives limited aquifer support and is developed by waterflood. There are 32 well slots of which currently 15 are for producers with a further 10 wells dedicated to water injection. From the outset it was recognised by BP that the formation water / injection water mix would lead to a severe scaling tendency[1]. The current operator, Talisman, have sought to review the scale management process to ensure that any lessons that can be learned from analysis of the earlier stages of production may be applied to ensuring effective scale control to the end of the field life cycle. It is the results of that review process that are presented in this paper. Reservoir Description and Field Development Gyda hydrocarbon reserves are contained in Upper Jurassic shallow marine sands. Reservoir depth is 3,650 - 4180 m (11,975 - 13,665 ft) subsea, initial temperature was 160 °C at 4,155 m (320 °F at 13,362 ft) and initial pressure was 604.5 bar at 4,155 m (8,768 psia at 13,362 ft). Some areas of the reservoir are heavily faulted, while others are moderately faulted. The sands are bioturbated, and in areas they are interbedded with calcite stringers. The field is divided into three regions, main field, the South-West and Gyda South with different PVT regions. The main field has a dip-closure in the western parts, called the C-sand area. The crest area has closure in the east by the Hidra fault system while it pinches out to the south. The downdip area has closure to the south by a Triassic high, while the southern area (Gyda South) is a tight rollover on the western bounding fault of the horst block. The field is moderately faulted (Fig. 1), and production has demonstrated that communication in some reservoir layers is good. Nevertheless, several distinct field compartments are defined from pressure data, and geochemical data from Gyda South indicate that sealing faults controlled the filling and cementation history[1]

Comparison of Streamline Reservoir Simulation of Brine Flow, Mixing and Barium Sulphate Induced Formation Damage with Observed Produced Water Chemical Data to Aid Scale Management

“…1 shows that there is a progressive decline in [Ba] (barium concentration) at low IWF (0%~50%), then [Ba] can be observed to remain very low in the produced brine from 60% IWF and above. As a whole, this kind of Ba trend is similar to some typical barium behaviours which have been frequently observed in other seawater flooded North Sea sandstone reservoirs (Paulo et al, 2001;Mackay, E.J., 2006;Houston et al, 2006;Gomes et al, 2012). However, in the Miller field, the depletion in Ba at all seawater fractions is significantly lower than observed previous in other fields.…”

“…However, the role of the reservoir was totally neglected in all of thermodynamic modelling work which is based on the fundamental assumption that the potential scale precipitation will be due to pure mixing of formation and injected water in the production well. However, it has been reported in some recent publications that both of brine mixing and geochemical reactions take place within the reservoir and the chemical composition of produced water would then be altered by them before arriving at producers (Paulo et al, 2001;McCartney et al, 2005;Houston et al, 2006;Mackay et al, 2006;Gomes et al, 2012;Fu et al, 2012). This paper reveals some fluid/fluid and fluid/rock interactions occurring in the reservoir through combining produced water chemical compositional data with geochemical and reservoir simulation model.…”

Impact of Reservoir Reactions on Thermodynamic Scale Predictions

“…Therefore, reservoirs E, F, G and H, as well as the Gyda field 6,12 show strong evidence that sulphate reacts with calcium deep in the reservoir. It is important to observe that in all the cases the reservoir temperature is above 120ºC and the calcium concentrations in the formation water are higher than 7000 mg/l.…”

“…In this case sulphate stripping in situ was observed due to precipitation of calcium sulphate 6 . Indeed, the Gyda field is not an exception 13 and the work presented in the current paper has identified that in reservoirs with temperatures above 120ºC and rich in calcium (> 7000mg/l) the typical behaviour is for anhydrite precipitation deep in the reservoir.…”

Re-development of the Frøy Field: Selection of the Injection Water

“…Several studies have demonstrated that reactions occurring in the reservoir as a result of waterflood can result in modifications to the produced water compositions and scaling potential (Mackay et al, 2006;McCartney et al, 2010a;McCartney et al, 2007;e.g. Sorbie and Mackay, 2000).…”