SPE Production & Operations
Volume 22, Number 4, November 2007, pp. 434-441

SPE-100520-PA

Placement Using Viscosified Non-Newtonian Scale Inhibitor Slugs: The Effect of Shear Thinning

View full textPDF ( 1,506 KB )

DOI  More information 10.2118/100520-PA http://dx.doi.org/10.2118/100520-PA

Citation

  • Sorbie, K.S., Mackay, E.J., Collins, I.R., and Wat, R. 2007. Placement Using Viscosified Non-Newtonian Scale Inhibitor Slugs: The Effect of Shear Thinning. SPE Prod & Oper22 (4): 434-441. SPE-100520-PA.

Discipline Categories

  • 5 Production and Operations
  • 5.3.5 Scale, Sand, Corrosion, and Clay Migration Control
  • 5.3.4 Acidizing
  • 5.5.1 Asphaltenes, Hydrates, Precipitates, Scale, Waxes (Inhibition and Remediation)

Summary

Reservoir formations are often very heterogeneous and fluid flow is strongly determined by their permeability structure. Thus, when a scale inhibitor (SI) slug is injected into the formation in a squeeze treatment, fluid placement is an important issue. To design successful squeeze treatments, we wish to control where the fluid package is placed in the near-well reservoir formation. In recent work (Sorbie and Mackay 2005), we went “back to basics” on the issue of viscous SI slug placement. That is, we re-derived the analytical expressions that describe placement in linear and radial layered systems for unit mobility and viscous fluids. Although these equations are quite well known, we applied them in a novel manner to describe scale inhibitor placement. We also demonstrated the implications of these equations on how we should analyze placement both in the laboratory and by numerical modeling before we apply a scale inhibitor squeeze. An analysis of viscosified SI applications for linear and radial systems was presented both with and without crossflow between the reservoir layers.

In this previous work, we assumed that the fluid being used to viscosify the SI slug was Newtonian(Sorbie and Mackay 2005). However, the question has been raised concerning what the effect would be if a non-Newtonian fluid was used instead. We mainly consider the effect of shear thinning, although our analysis is generally applicable if the non-Newtonian flow rate and effective viscosity function is known. We address the questions: Does the shear thinning behavior result in more placements into the higher or lower permeability layer (in addition to the effect of simple viscosification)? Can the shear thinning effect be used to design improved squeeze treatment?

Background and Introduction

Chemical SIs have long been applied in downhole “squeeze” treatments to prevent mineral scale formation(Miles 1970; Vetter 1973; Meyers et al 1985; King and Warden 1989; Yuan et al. 1993; Boreng et al 1994; Sorbie et al. 1994). In a homogeneous reservoir layer, adsorption may be the only retention mechanism governing the SI return from the well. However, reservoir formations are rarely homogeneous but are made up of highly heterogeneous rocks which may have a layered or more complex structure as determined by various sedimentological, structural, and diagenetic factors(Weber 1982). Here we will consider only layered systems where the various layers have different permeabilities, k (and porosities, f) in the near-well formation. In such systems, SI placement within the formation is an additional aspect of a squeeze treatment that must be considered because this may affect the SI returns.

Scale inhibitors are typically applied as aqueous solutions at concentrations, in the range 10,000 to 150,000 ppm. These solutions usually have a viscosity (m) close to that of an injection brine; (i.e., ~1 cP at 20oC and 0.3 cP at 100oC). Therefore, apart from a slight temperature effect, the injected brine displaces formation water (FW) at unit mobility. Also, for lighter oils, a unit mobility displacement is often involved although viscosity and relative permeability effects may be more important in heavier oils. In unit mobility injection into a heterogeneous layered linear or radial system, as shown schematically in Fig. 1, the fluid placement into layer i is governed solely by the (kh)i product. That is, injecting fluid at a total volumetric flow rate of QT into an N-layer system of the type shown in Fig. 1, then flow into layer i, Qi, is given by:

Figure (1)

It can easily be shown that this is true for unit mobility displacement in a linear or a radial system with or without crossflow. However, this well established result might foster the belief that linear and radial systems are also very similar under viscous slug injection with and without crossflow and this is not the case.

In recent years, the use of viscosified slugs of SI has been proposed to change the placement pattern in a “favorable” manner (Mackay et al. 1998; Feasey et al. 2004; Mackay and Al-Mayahi 2003; Jordan et al. 1999). In this context, “favorable” may mean to place the SI slug entirely in the high permeability layer from which the water is being produced. However, it may also mean that we wish to place the inhibitor slug in the lower permeability layers where it may be “stored” and flow back to the well more slowly because of the reduced flows from these layers. Whatever our intention, we must clearly understand the fluid mechanics of viscous slug placement in heterogeneous systems to achieve the effects we are after (i.e., most of the SI slug in the high k or low k layer).

At this point, we note that viscosified solutions or other types of “diverter” may also be injected to modify the relative flows in the wellbore and near-well formation in long horizontal wells(Mackay et al. 1998; Feasey et al. 2004; Mackay and Al-Mayahi 2003; Jordan et al. 1999). Viscous fluids are also used in a similar manner in viscous acidizing where the central intention is to “present the treatment fluid evenly to the face of the formation”. However, we will not consider intra-wellbore effects in this paper. We are primarily concerned with the fluid mechanics in layered heterogeneous formations both with and without crossflow for linear and radial systems. These layered systems may have N-layers but for simplicity, we consider only 2, as in the simple schematics in Fig.1.
Most results here can be generalized quite easily to multi-layer systems. In our previous paper, we described both analytical and numerical result for viscous SI placement using a Newtonian fluid (Sorbie and Mackay 2005). Here, we generalize these results to where a non-Newtonian (shear-thinning) fluid is used to viscosify the SI slug to “help” in the placement.

View full textPDF ( 1,506 KB )

History

  • Original manuscript received: 16 March 2006
  • Meeting paper published: 30 May 2006
  • Revised manuscript received: 20 December 2006
  • Manuscript approved: 5 January 2007
  • Version of record: 20 November 2007