|
Safety and Integrity
|
|
Alarm Management
|
|
Process Engineering
|
|
Remote Monitoring
|
|
Hydrocarbon Accounting
|
|
Pipeline Leak Detection
|
|
Production Data Management
|
|
Environmental Monitoring
|
|
Remote Monitoring  We have developed a technique for identifying and analysing poor process and control performance using information already available in most plant control systems.
Process plants operate best under steady conditions. Yet many operations experience periods of unsteady operation that are not severe enough to cause a process upset or trip, but have an impact on performance nevertheless. The results of our technique can be monitored remotely, using a standard browser and the company Intranet, or even the public Internet.
Causes of unsteady operation
The most common causes of unsteady operation include:
- external influences (unsteady feed, surging oil well, pipeline slugging);
- instrumentation faults (control valve ‘stiction’, hysteresis, and slew rate limitation, noisy transmitter signal);
- controller performance (sub-optimal or inappropriate PID settings, inadequate filtering of noisy measurements, integral wind-up);
- process and operational problems (interactions, operator intervention).
Impact on the business
Acute process instability, upsets and trips have an obvious effect on business performance, but are usually dealt with promptly. On the other hand, the plant operators tend simply to cope with (or not even notice) the less severe disturbances. These then become chronic, part of the ‘norm’.
Typical consequences include:
- poor/variable oil-in-water results for produced water discharges from offshore installations;
- production constraints, particularly related to compressors, pumps and other machinery;
- 'giveaway’ where product quality is kept well above specification in order to allow for the occasional poor result;
- production loss where external disturbances regularly cause trips owing to lack of control robustness.
Finding and fixing
In principle, most operations have the technical ability to identify unsteady and underperforming plant, and to track down and rectify the root causes. However, in practice, this is often not done owing to cost and resource constraints, and pressure of other work. Offshore installations are particularly vulnerable to this, as are other remote unmanned, or minimally manned sites.
Our remote monitoring and diagnostic technology identifies areas of the plant that are under-performing and helps the client’s staff to determine the root causes.
Performance measures
Our technology uses three measures of process stability and controller performance:
- time-weighted average deviation from set point (for controller PVs) or from medium-term average value (any analogue value);
- time-weighted average rate-of-change (any analogue value including controller PVs and controller outputs);
- time-weighted average amplitude (any analogue value including controller PVs and controller outputs).
Implementation
The algorithms used to derive these measures can be implemented using the calculation and dynamic functions provided by most digital plant control systems and SCADA software packages. The results can be monitored using a ‘traffic lights’ type of display (green = normal, yellow = deteriorating, red = unacceptable) and can be analysed using trend graphs.
Data compression
As these measures relate to stability and performance in the medium to long term, the results need only be transmitted to the remote monitoring location at a relatively low frequency, typically every 1 – 10 minutes. The data compression implicit in the algorithms is useful where bandwidth is restricted.
System requirements
Real Time's stability and performance monitoring algorithms can be installed on any plant control or real-time monitoring system that supports basic arithmetic and dynamic operations. For remote monitoring, a means of transmitting the algorithm results for display, historisation and trending is also required. (The example on this sheet was created using the DeltaV distributed control system and DeltaV Web Server.)
|