The prediction of pH-dependent scales such as carbonates and sulfides presents unique challenges because their formation is strongly related to the three phase partitioning of the acid gases (CO2 and H2S). A rigorous procedure is required to ensure proper modelling of the hydrocarbon phases, in order to derive the correct data input for the software from available field data. Using this input, reliable scale prediction calculations may then be run using either integrated or separate PVT and scale prediction software. Although some carbonate scale prediction methods have been published in the past, these methods are field and software specific, and they do not provide a general procedure for carbonate and sulfide scale predictions in oil and gas wells. Operators also have in-house proprietary procedures, but these are not publicly available and hence cannot be used or critically reviewed by the wider upstream chemistry community.
This work presents an improved version of the original Heriot-Watt scale prediction workflow previously published in 2017 (Verri et al., 2017). The authors show how this previous procedure is intrinsically general, but requires modifications depending on specific field variables such as oil type, topside vs reservoir data availability or the use of integrated or separate PVT and scale prediction software.
The workflow is built on three general calculation blocks which apply to all field scenarios, as follows: 1. defining a total PVT feed; 2. modelling the water chemistry leaving the reservoir; 3. running scale prediction calculations throughout the system. After describing the general carbonate and sulfide scale prediction procedure in details, this paper also looks into the specific calculation steps required in different scenarios for variable oil type, sample availability (topside vs downhole), software choice (integrated vs separate PVT and aqueous phase models), EOR, reservoir souring, artificial lift, and HP/HT/HS.
This is a truly general approach to carbonate and sulfide scale predictions which the authors hope will provide a widely available, useful tool to anyone performing field prediction studies for pH-dependent scales. In addition, a worked example is presented in Appendix 1 where the "correct" answer is known, and the workflow is demonstrated to be rigorous in that it recovers this correct answer. This data can be used to test any other procedures for rigorously predicting carbonate and sulfide scales from field data.