Investigation of SAGD Steam Trap Control in Two and Three Dimensions

Paper Number: 50413-MS

Authors: Neil R. Edmunds, Clearwater Engineering

Source: SPE International Conference on Horizontal Well Technology, 1-4 November 1998, Calgary, Alberta, Canada

Copyright: 1998. Society of Petroleum Engineers

Steam trap production control, i.e. producing fluids to maintain BHT just below the boiling point of water, has been traditionally recommended for SAGD operations. In this work a numerical investigation of the relationships between producing steam trap subcool setting and related parameters of interest such as fluid level, pressure, production rate, and steam/oil ratio is reported for a prototype Athabasca reservoir.

It is found that steam trap dynamics are rather more complex than previously imagined. In three dimensions and the real world, for example, inflow temperature can and does vary significantly along the well according to local conditions, resulting in 3D simulations predicting significantly lower production rates than equivalent 2D cases. The findings have implications to many aspects of SAGD engineering including process optimization, forecasting, completion design and installation, production operations and performance analysis.

Optimum SAGD requires that the steam chamber be kept well drained, so that liquid does not accumulate over the producer but neither is steam produced. According to Butler, the analogy of this fundamental operation to that of an industrial steam trap was part of the earliest concept, prior to any testing or calculation:

"It was thought that if [oil and condensate] were not removed too quickly, then the tendency of the steam to flow directly to the production well...could be reduced or possibly even eliminated. This is analogous to the ability of a steam trap to allow the flow of condensate from the bottom of a steam-heated radiator without allowing significant bypassing of steam."

Since then (1978), many eloquent rehashings of the function and necessity of proper production control have been published. The literature is however thin on specifics of implementation and operation, regardless of whether the subject was an analysis, a numerical study, a physical model, or a field test.

Real steam traps are used for steam heating control in refineries, and other process plants. They are mainly of two types, float and thermodynamic. The float type operates directly on the condensate-steam interface, letting liquid out but retaining vapor, somewhat like a toilet float. The thermodynamic type sits downstream and below the actual interface, and merely infers its nearness by comparing the local temperature and pressure with the steam saturation curve. The proper temperature setting, and often the provision of a heat sink by removal of insulation from the condensate leg, are critical for proper and stable operation of a thermodynamic trap.

In Butler's analytical models, the producing boundary condition is implicit in the geometrical assumption that the steam zone is anchored at the production well. A production control method has to somehow ensure that this assumption is the operating reality.

In scaled physical models, production control has largely been done manually based on visual observation of the actual steam level. In the case of atmospheric models, simple gravity has sufficed to drain liquids. In one case, a production line temperature was monitored to assist in steady control: Joshi and Trelkeld reported production temperatures about 20°C below steam temperatures were generally sufficient to establish a definite liquid leg above the producer, with no short circuiting of steam. The model was operated at slightly above atmospheric pressure, however, and was not well-scaled thermally.

Close Window