Natural gas processing

Using process simulation for sizing and construction scheduling of gas gathering systems

Designing gas gathering networks can be a challenging task for engineers due to several factors, including multiphase flow (presence of condensate) in the network, uncertainty in forecasting production data and exact pipeline routes, looping/networking of pipelines connecting the wells, as well as the location of multiple processing facilities. In this article, we demonstrate how process simulation was used to predict the required pipeline sizes in different segments of a gas gathering network and overcome these challenges. This information is crucial for budgeting and construction scheduling. In this case study, the gathering system includes 32 operating wells, for which the production forecasting data was provided for each quarter from 2014 to 2022. Figure 1 shows a snapshot of the gas gathering network at the end of 2022.


Figure 1. Schematic of gas gathering network


The model was set up in Aspen HYSYS Hydraulics. HYSYS Hydraulics is a pipeline network modelling platform which allows for complex gathering systems to be modelled, allowing for pipeline looping, multiple “sinks” (plant inlets), etc. Rigorous multiphase pressure drop and heat transfer correlations are incorporated. An automation tool was developed by Process Ecology to transfer the forecast data to the simulation model and to report and interpret the simulation results to identify segments of the network where constraints were violated. Integration of the tool with Google Earth enabled rapid estimation of unknown pipeline profiles.


Figure 2. Elevation profile of Well 4 to Well 5 pipe from Google Earth


As shown in Figure 1, there are two sinks in the model: Plant A and Plant B. It is important to note that due to limited capacity, Plant A is only capable of handling 50 MMSCFD of gas; the rest of the production from Wells 1-9 would be directed to Plant B. The Plant B suction pressure was set to 1200 kPag while Plant A suction pressure was calculated based on the modeling results so that only 50 MMSCFD of gas was sent to Plant A. The network was designed so that the ratio of velocity to erosional velocity would not exceed 55% for a specific period of time.

Based on current operating data, the model was calibrated and the appropriate pressure drop correlation was selected.  Provided forecasting data indicated a fairly liquid-rich gas with 25-35 bbl/MMSCF of liquids in the produced gas. Therefore, selection of the appropriate pressure drop correlation was crucial. The calibration was done based on the provided data for Well 9 for which the flow was directly routed to Plant A and could be easily isolated from rest of the network. Figure 3 shows the results using the Tulsa pressure drop correlation; the model predicts pressure drops within 20% of field measurements.


Figure 3. Model vs. field data - pressure drop for Well 9


The developed model coupled with the Process Ecology automation tool enabled quick analysis of the system at each quarter allowing for identification of the network segments where velocity would exceed 55% of erosional velocity. Table 1 shows a sample of output results for two segments of the network. As shown in the table, the 12” pipe from Well 22 to Well 16 had been identified to violate the velocity constraints, which would indicate the need for installation of a new larger-diameter pipe.


Table 1. Sample of output results from the Process Ecology automation tool


Based on the provided forecasting, and through consultations with the client, it was decided to install a 16” pipe in this segment of network to alleviate the high velocity issue.

Besides providing critical data for scheduling and budgeting purposes, the developed automation tool allowed for notifications regarding technical issues, such as potential reversal of flow direction in certain segments of the network which would require updates to the pigging procedures. In this scenario, reversal of flow takes place in the Riser 1 to Plant A pipe segment, where initially the direction of flow is from Riser 1 to Plant A. However, the direction of flow changes in Q2-2018 due to higher production from Wells 1-9.


Conclusions

This article demonstrates how process simulation can be used to model gathering networks and provide the required data to improve long-term planning. The Process Ecology automation tool helps identify design issues, determine flow rates and well pressures, and recommend changes based on the results and sizing criteria of the network. Modelling provides valuable information for scheduling and budgeting purposes and helps to identify potential of reversal of flow direction in the network as well as other key design considerations.  


Do you have questions or comments regarding this article? Click here to contact us.


Search

Tags

Latest articles