Page 106 - Intelligent Digital Oil And Gas Fields

CHAPTER THREE

Data Filtering and Conditioning*

Contents

3.1 DOF System Data Validation and Management 76
3.1.1 Data Processing 78
3.2 Basic System for Cleansing, Filtering, Alerting, and Conditioning 79
3.2.1 Data Validation System Architecture 80
3.2.2 Advanced Validation Techniques 84
3.2.3 Model-Based Validation Methods 85
3.2.4 Data Replacement Techniques 85
3.2.5 Data Reconciliation 88
3.3 Conditioning 91
3.3.1 The Level of Rate Acquisition (Data Frequency) 91
3.3.2 Down Sampling Raw Data 93
3.3.3 Summarizing From Raw Data 96
3.3.4 Well and Equipment Status Detection Required for Sampling 97
3.4 Conclusions 99
References 99

All digital oil field (DOF) systems generate high-frequency data from mul-
tiple sensors from most sources in the field. These sensors communicate
through the SCADA systems, remote terminal units (RTU), and data his-
torians with multiple corporate data systems as described in Chapter 2.
It does not matter if the workflow is for surveillance, regulation,
reporting, optimization, or control; timely and accurate data is absolutely
required. However, the requirements for what constitutes “timely” and
“accurate” can vary widely for these workflows, depending on their nature
and urgency. For example, gas lift flow regulation requires sub-minute fre-
quency and highly precise indication of flow and valve position, while gas lift
optimization may require only the valve position and likely needs only
hourly or daily indications of gas lift volume and oil production volume.
Thus, DOF workflows have differing requirements for high-frequency data
and some may use lower frequency data. The data may be acquired at

*With contributions from Doug Johnson.

Intelligent Digital Oil and Gas Fields © 2018 Elsevier Inc. 75
https://doi.org/10.1016/B978-0-12-804642-5.00003-7 All rights reserved.