The data from online monitoring serves as a crucial basis for assessing the operational status of three-stage oil and gas recovery. However, the data itself can also be inaccurate due to factors such as sensor drift, improper sampling point locations, and conversion methods that do not align with reality. All these can lead to seemingly normal readings masking real issues. Therefore, it is important not to rely solely on the numerical values of the monitoring data, but also to regularly verify them against the actual on-site conditions. The purpose of verification is not to prove the accuracy of the data, but to detect any inaccuracies as early as possible.
Why should we conduct data verification
The readings of the monitoring system come from sensors and conversion models, both of which may deviate from reality. Sensors may experience zero drift after long-term operation, and if the sampling point is set in a position where the airflow is unstable or prone to fluid accumulation, the readings will be out of touch with the actual working conditions. If the conversion method is not adjusted and updated along with the process, the results obtained will also deviate systematically. These problems will not make the data appear bad, but will quietly make the data inaccurate.
Data sources and comparison methods
The basic idea of verification is to compare two types of data from independent sources: one is the continuous readings from online monitoring, and the other is the results obtained from on-site manual measurements or material records. Since the two types of data originate from different sources and have different sources of error, if they are close to each other, it indicates that the system is basically reliable. If they continue to deviate, it indicates that at least one of them needs to be investigated.
Three steps for verification implementation
- Determine the verification cycle and sampling time points, and try to select periods with relatively stable operating conditions, avoiding the drastic fluctuations caused by concentrated operations;
- Synchronously record online readings and on-site measurements, specifying the measurement method, instrument, and the operating conditions at that time;
- Record several consecutive cycles, first observe the trend and then examine individual points, avoiding drawing conclusions based on a single deviation.
How to judge deviation
To judge deviations, it is important to consider whether the direction and magnitude are stable, rather than just how much a single reading differs. Stable small deviations usually stem from the conversion method or the sampling location itself; deviations with continuously expanding magnitudes often indicate changes in the sensor or equipment status; deviations with unclear directions and varying magnitudes are more likely caused by operating conditions or operations.
| Deviation manifestation | Possible sources | troubleshooting direction |
|---|---|---|
| Stable fixed bias | The conversion method or sampling location is inappropriate | Check the conversion method and sampling point location |
| The deviation is gradually widening | Sensor drift or component status change | Calibrate the sensor and check the operating status of related components |
| Irregular fluctuations | Changes in operating conditions or operational differences | Record and compare homework sessions |
| Multiple points experiencing simultaneous anomalies | Systemic issues or power signal interference | Check the power supply and signal circuits |
The approach to handling deviations once discovered
First, check the sampling and conversion, then the sensor, and only then suspect the equipment itself. The reason for this order is simple: the cost of troubleshooting the first two items is low and the probability of their occurrence is high. However, once you directly check the equipment, it is easy to introduce new variables during the disassembly and assembly process. The troubleshooting process should be recorded, otherwise, when the same deviation occurs again, you will have to start from scratch.
Frequently Asked Questions
How often is it appropriate to conduct a verification?
Online data can be recorded daily or by shift, and manual verification is recommended to be conducted monthly or quarterly. An additional verification should be performed after process adjustments or equipment maintenance.
If the online data and on-site measurements are inconsistent, which one should prevail?
The party with reliable methods and traceable processes shall prevail, and any discrepancies shall be recorded. In most cases, on-site simultaneous measurements can better reflect the actual working conditions at that time, but this is only possible if the measurement methods and instruments themselves are reliable.
Does the verification require a production halt?
Generally, it is not necessary. Verification is mostly completed under normal operating conditions. It is only necessary to select a period with relatively stable operating conditions and record the operating status at that time.
This article is compiled by Luoyang Wohong Petrochemical Equipment Co., Ltd. The company provides design and manufacturing, skid-mounted customization, installation and commissioning, as well as after-sales operation and maintenance services for three-time oil and gas recovery equipment to gas stations and oil depots. If you need on-site inspection or scheme configuration, please feel free to contact us for discussion.