Well’s Liquid Production Rate Prediction Method Utilizing Temperature Profile Reverse Calculation, One Step Closer to Virtual Production Test on Limited Wellhead Data

Abstract

Wellhead data and the actual production rate measurement are the most parameters evaluated to be prepared for any production dynamic by still respecting to the downstream facility limitation. One of the them is liquid production rate. Case study discussed is Tunu Field in Mahakam. Although it is a gas field, liquid rate is important information to shut in high liquid producer wells prioritization in case of limitation on the production facility (i.e liquid transfer pump break down).

Production parameter is nearly blind on well level. Flowing pressure, temperature, and gas rate by orifice flowmeter are locally updated by weekly visit. While well’s liquid production rate refers to periodical collective well test from test separator or clean up history by mobile testing unit. Unfortunately, 73% from 29 test separators were unfit to measure liquid due to aging. While mobile testing units were limited and costly for routine well production test.

Well’s characteristics vary depend on the open reservoir. From average 220 active wells, most of them were strong water drive on shallow reservoir and depletion drive gas or retrograde condensate with water contact in the deeper reservoir. Therefore, liquid production was inevitable on the late production phase. With the absent of updated well test data, liquid producer well  identification became very challenging. In 2021, liquid estimation accuracy on field level was 50% compared to actual liquid production. At this period, it was estimated 961 MMscf production lost on recorded emergency events due to false identification or ineffective decision in shut in high liquid wells.

The initiative came up with development of virtual well’s liquid production prediction. The concept is reverse calculation of temperature profile estimation on flowing fluid inside conduit. The overall heat transfer coefficient (u-value) as one of critical parameter was categorized based well’s completion type by referring to more than 1500 historical test data. It was combined with wellhead parameter and reservoir data to calculate estimated liquid. Using combination of nodal analysis software and VBA based
spreadsheet, an in-house tool was developed to provide quick and consistent calculation on all of active wells. On the field level, the accuracy of liquid prediction was significantly increase to 90 – 110 % and all high liquid producers could be well identified in 15 minutes.

This initiative provided a quick and adequate well’s liquid production estimation at free maintenance and operation cost to be ready at any time, one step closer to virtual well production test for wider application.