Producer‐flowline temperatures (FLTs) can be measured automatically with a thermistor on an emergency‐shutdown system (ESD), or manually on a specified spot on flowline with a handheld unit. Measured FLTs can usually be mapped to represent the formation‐temperature distribution for steamflood reservoir management purposes (Hong 1994; Nath et al. 2007). In addition to FLT, wellhead temperature (WHT) is another surface temperature. Predicting the long‐term WHT trend in steamflood operation is necessary for designing surface facilities for both oil dehydration/separation and produced‐water recycling. This predicted temperature will also be applicable for production‐performance monitoring.
To predict the wellhead temperature, Hasan et al. (2009) derived a steady‐state analytical solution for calculating WHT from bottomhole temperature (BHT) under flowing conditions of a multisection slant wellbore for the isothermal primary‐depletion process, with both WHT and BHT being time independent for a given gross rate (Igec et al. 2010). This steady‐state analytical solution has been extended to calculate steamflood producer WHT from BHT (both are time dependent) by consecutively approximating the WHT and monthly average of FLT measurements to a steady‐state solution. The monthly averaged FLTs are seasonally variable and higher in the summer months of July through September and lower in the winter months of December through February. Monthly averaged FLT measurements depend on an annual ambient‐temperature cycle within the depth needed for reaching an undisturbed ground temperature (typically 30 to 50 ft) (Gwadera et al. 2017). WHT, if measured, should be comparable with FLT for their close typical distance of 5 to 10 ft. WHT prediction, however, is only process dependent and not seasonally variable because of the inability to describe seasonally undisturbed depth in the geothermal gradient. Therefore, WHT prediction can be validated with average summer‐month FLT measurements when heat loss becomes minimal. BHTs in this analytical approach are predicted by the Lauwerier (1955) analytical model and improved by calibration with the available reservoir‐simulation model or several years of FLT measurements for steamflood response time.
The objective of this study is to develop an integrated production‐monitoring approach using only the surface information, including WHT and FLT, oil/water‐production rate, and injection‐pressure/rate data, which can be applied to diagnose and optimize steamflood production performance. A field case study for the South Belridge Diatomite steamflood was investigated. WHT prediction is compared with FLT measurement for diagnosing and understanding the production performances, such as premature water or steam breakthrough, interference by the waterflood on the steamflood boundary producers, as well as the FLT variation related to the target rates for steam injection. This diagnostic analysis approach combined with the Buckley‐Leverett theory‐based displacement‐efficiency analysis, and injection pressure and rate signal, will help to develop an improved understanding of the displacement detail and form a decision base to optimize the production performance.