Abstract
In today's Oil and Gas Industry, as we evolve more towards smart fields and integrated operations, the term digital oilfield is being used more often. Many companies implement smart fields where transmitters and gauges are installed in the field along with a Supervisory Control and Data Acquisition (SCADA) system. One common threat faced in a digital oilfield setup is the failure of accurate data transmission. This includes no data transmitted, erroneous/wrong data transmitted, error in the calibration of the field devices, network interruptions, failure of SCADA system etc.
This paper describes the development of a methodology used to overcome the issues of bad data from the field with the usage of data conditioning and also the suitable data tolerance which can be accepted before the system fails. This method uses smart rules and automated data conditioning which involves historical data, data from nearby wells and results from well modeling workflows. When applied in the field, it ensures that the system runs even when not all the required data from the field is acquired for successful workflow execution.
Implementing data conditioning rules described in the paper improved the accuracy of the workflow from 71% to 98% which represents a huge improvement. The solution was applied in an offshore field. The automated workflows involved in this solution was well status identification, real time production surveillance, real time gas lift optimization, theoretical production rates from well models, well test validation and platform monitoring. As more data becomes unreliable and missing, the confidence level of the results of the workflow also decreases to a certain point where too many required data points are missing or not available and the system would send an alarm and flag the results of the workflow due to unreliable data transmission. This scenario will be seen if there is a complete failure of the SCADA system or severe network interruptions for an extended period of time. The solution and data conditioning methods applied managed to increase production by reducing the amount of time needed to identify wells which quit offshore which reduces deferred production.
The paper provides a quantitative assessment of the benefits realized by applying the data conditioning and using suitable data tolerance when running workflows used for production surveillance and optimization in a digital oilfield. As we try to achieve higher production rates and greater recovery from the field, more high frequency data is required and the methods described in this paper can be applied to other fields to solve the missing and erroneous data transmission problems.
