Abstract

The flow behaviour of cement slurries, as measured with different types of equipment - a pilot-scale pipe viscometer, a Searl and a Couette coaxial cylinder viscometer - is analyzed.

The dimension of the flow geometries, respectively pipe diameter and annular gap size, was varied to detect the presence of an apparent slippage at the wall, Different cement formulations were tested including a wide range of specific gravity from 1.7 to 2.2. Experimental flow slurries were also compared for a polymer solution, and two water-base bentonitic muds.

For cement slurries. experimental results show that the flow curves obtained with the pipe and the Searl viscometer are very similar. The abrupt increase in viscosity at low shear- rates is attributed to the existence of a yield stress (Tv). For stresses lower than T,, flow is still observed in both cases. The size dependence of the flow curves indicates that such a behaviour can be attributed to an apparent slippage at the wall. For stresses higher than tv, little or no slippage is observable, and slurries exhibit a shear thinning behaviour that can be well fitted to a power law model from 30 to 500s.

Different methods of calculating pressure drops in pipe flow from rotational viscometer data are presented. Their precision and limitations are compared. As a conclusion, it is established that laminar pipe flow pressure losses can reliably be predicted from rotational viscometer measurements if made on representative samples using equipment which minimizes particle centrifugation e.g. Searl systems instead of Couette systems.

Introduction

In the oil industry, the current practice for the rheological characterization of fluids is based on rotational viscometry The API I recommends the use of a coaxial cylinder viscometer, along with a very precise procedure for testing cement slurries, and indeed this technique gives quite repruducible results (table 1). However the physical significance of the results obtained has been questioned because the measurements were shown to be dependent on the gap size 121. A diameter dependency of the How curves was also found in pipe flow experiments. These phenomena were attributed to an "apparent slippage- at the wall, and they will be discussed in details for both pipe and coaxial flow of different cement and mud formulations. As far as coaxial flow is concerned, our experience is that particle migration due to centrifugal forces is by far the predominant factor affecting the rheological measurements on cement slurries. A reduction in the centrifugal forces. and therefore more reliable measurements, are obtained when rotating the inner cylinder - Searl viscometer illstead of the outer one - Couette viscometer - everything else being equal, The objective of the present paper is to validate this approach by comparing measured friction pressures in pipes, with the calculated ones from Searl and Couette viscometer data.

Experimental investigation

The rheological characterization of the tested fluids was performed using two rotational viscometers - a Couette and a Searl - and a pilot-scale flow loop. The Couette system is the Fann. The tests are run with geometry RIB1 (table 2) and spring 0.5, according to API specification 10.

The Searl system is also a commercially available equipment- the CS100 from Carrimed. It has the peculiarity of being operated in constant torque mode instead of the usual constant speed mode. The torque can be applied with an accuracy of +/.01% between 10-5 and 10-2 N.m.

P. 755^

This content is only available via PDF.
You can access this article if you purchase or spend a download.