Within the digital or business transformation stream of the Emerson Exchange EMEA 2024 conference, Jacek Brauer, Europe and Asia advanced process control expert at Bunge, explained how to control non-linear processes efficiently using DeltaV™ Distributed Control System function blocks.
Bunge is the world leader in oilseed processing and a leading producer of specialty plant-based oils, fats and protein. To ensure protection of proprietary information Jacek avoided discussing process descriptions, parameters, absolute values, units and measurement ranges, instead focusing on DeltaV system trends, settings and process dynamics as a gain, time lag and time constant.
Process data quality is critically important for successful analytics. The first step is to perform tuning of basic control loops. The basic principle of good control loop turning is “squeeze and shift”. Squeeze is used to describe improving control quality and performance, i.e. reducing process variability, while Shift relates to the moving of process parameters closer to their limits.
Heating a product using steam is a very common practice within the process industry, but there are several challenges, such as non-linearity of the process and higher-order inertia. It is not easy to determine process gain, which will vary depending on the plant capacity. There are also different characteristics during thermal power increases and decreases, as well as time lags. Common issues include poor control quality, large deviations and oscillations that go beyond limits, and slow response to disturbances. As a result, there is often suboptimal energy consumption, an impact on useful heater lifespan (thermal stresses) and degraded quality of thermal sensitive media. Low temperatures often cause heat exchanger flooding.
Brauer explained that a simple application of PID control application cannot meet expectations and usually process parameters are far from optimal values, thus creating considerable energy losses. He presented several different control structures, including a very simple structure, cascade, cascade ratio, advanced cascade/feed forward and simplified feed forward. A cascade structure is used to separate the dynamics of fast changing steam pressure and slow changing product temperature.
Perfect control with adaptive feed forward
The DeltaV Saturated Steam Properties – Given Temperature (SST) function block calculates steam enthalpy, entropy, specific volume, and pressure for saturation conditions specified by a given temperature. Utilizing the DeltaV SST function block means there is no need to determine PID gain. The PID/gain is always correct/accurate, regardless of the process operating point – the nonlinearity problem is therefore solved. When implemented, the DeltaV function block enabled the process response to be a matter of seconds. This creates the possibility for almost perfect adaptive feed forward control. When applied to a 1.4 MW highly non-linear heating process, better regulation accuracy is achieved. Deviation is reduced from 4°C down to 0.1°C. An SP reduction of 2°C results in a 5% energy saving and required very minimal engineering effort and no control loop tuning.
