Update temperature control for a Claus unit with neural networks
Abstract
The Natref Oil Refinery in Sasolburg, South Africa, experienced refractory problems in the combustion chamber of the Sulphur Recovery Unit (SRU) since start-up (Figure 1). Even though the SRU was at all times operated within the temperature range of 1250 - 1750°C (as measured on the original thermocouples and the pyrometer), they experienced refractory failures, which can only be accounted for by temperatures in excess of 1850°C. It was concluded that the available temperature indication was not an accurate reflection of the actual temperature in the combustion chamber. Three new sets of temperature measurement were installed in order to better understand the temperature profile in the combustion chamber (each set consists of two thermocouples and one pyrometer). Furthermore, An ASPEN-based simulation was used to set up a neural network to predict the combustion chamber temperature based on three input variables, namely the oxygen enrichment level, the ratio of sour water stripper offgas to acid gas, as well as the amount of acid gas bypassed to the second chamber of the reaction furnace.
The reaction furnace was found to be operated at an average temperature of 1400°C during normal operation. This temperature is substantially higher than the normal recommended 1250°C for Claus combustion chambers. The neural network was connected to the Natref DCS and successfully proved that during the start-up of the SRU in 2006 a temperature in excess of 2000°C was reached inside the combustion chamber, explaining the refractory failure and subsequent shutdown of the unit.
In order to establish which temperature measurement should be used for control purposes during normal operation, the 9 temperature measurements over the length of the combustion chamber were compared to the neural network temperatures. All the other thermocouples (T1 - T6) were found to underpredict the ASPEN simulated temperature. Furthermore, the response of these thermocouples was much slower than that of the pyrometers.
With the exception of the burner box pyrometer, P1, all the pyrometers also underpredicted the ASPEN simulated temperature. The response of the pyrometers was, however, much faster and showed the strongest linear correlation with regards to the adiabatic flame temperature. The reason for the underprediction can be ascribed to heat losses in the combustion chamber. The middle pyrometer, P2, gave the smallest deviation between measured and predicted temperature values of only 28°C on average and should be used as the main control point with regards to temperature.
URI
http://hdl.handle.net/10394/13899https://www.hydrocarbonprocessing.com/magazine/2013/november-2013/cyber-security-and-process-control/update-temperature-control-for-a-claus-unit-with-neural-networks
Collections
- Faculty of Engineering [1123]