The shortage of fresh water resources in dry climate zones has increased the need to obtain drinking water or water for irrigation by desalination of sea water in regions close to the coast.
Currently around 1% of the world’s population depends on desalinated water for daily use, but this fraction is expected to grow rapidly . In principle, there are two main alternatives for seawater desalination: thermal processes (vacuum distillation) and membrane processes.
Seawater desalination is generally a very energy intensive process, but pressure driven membrane processes are typically less energy intensive than thermal processes . For seawater desalination, based on reverse osmosis, a semipermeable membrane is used to filter out the dissolved salt ions and obtain low-salt drinking water.
The salty water has to be pressed with high pressure through the membrane to overcome the osmotic pressure, a thermodynamic parameter driven by chemical potential differences of the solvent. This is achieved by using a high-pressure pump. The semipermeable membrane inside each membrane vessel restrains most of the salt. Only desalinated water (“permeate”) gets through the membrane, while the rest is rejected as “brine” with high salt concentration.
The high-pressure membrane unit is common to all reverse osmosis desalination plants although the details of pre- and post-processing may vary. To recover the energy content (high-pressure) of the salty water rejected by the membrane an energy recovery device is used.
It exploits the pressure content of the reject water to increase the pressure of fresh salty water coming from the ultra-filtration unit, which is afterwards further increased by the booster pump in order to reach the pressure in front of the membrane.
Pelton turbines were the first energy recovery devices deployed in reverse osmosis plants. However, nowadays mostly isobaric pressure exchangers (Fig. 1) are used because they provide higher energy efficiency. The use of instrumentation and control (I&C) technology in seawater treatment plants generally has the following objectives:
Sophisticated control engineering methods, which have become known as Advanced Process Control (APC) in other industry sectors, such as the chemical industry and oil refining industry, offer potential for optimisation of process control in the water industry as well.
Since these methods have been seamlessly integrated into modern distributed control systems such as SIMATIC PCS 7 and are made available at low cost as standard software blocks, nothing stands in the way of their successful application in desalination plants [3, 4]. Model Predictive Control (MPC) is particularly attractive in this context.
MPC allows “predictive operation” of the plant because it takes into consideration both the physical/chemical interactions between different variables and measurable disturbance effects.
MPC is integrated as a standard function block in the SIMATIC PCS 7 Advanced Process Library with the name ModPreCon.
The example to be examined is a membrane unit of a real-world seawater desalination plant in Arabia. A model is created in a Siemens internal tool for simulation of biological and chemical process technology. The membrane model is based on . Models of both high-pressure pump and booster pump use pump performance curves from pump datasheets.
Fouling is assumed to increase steadily in such a way that after 40 hours of unit operation, flushing of the membrane vessels is required. The simulation, when compared with measured data from the real plant, describes the essential dynamic processes of the seawater desalination plant effectively.
Plant operators need to increase the high pressure setpoint manually several times a day to compensate for fouling and to bring the permeate output back to an acceptable level. Each of these setpoint steps leads to a steep gradient in pump speeds, permeate flow and permeate salinity. Neither permeate flow nor permeate salinity is controlled directly.
Although the decline of permeate flow due to fouling is compensated manually, adjustments in permeate salinity cannot be performed by the operator at all. The design of a control concept starts from a hierarchical order of requirements for typical plant operation as defined by the company operating the plant:
|MV1 in bar HP pump pressure setpoint||MV2 in m3/h booster pump flow setpoint|
|CV1 in g/L permeate salinity||–||–|
|CV2 in m3/h permeate salinity||+||+|
|CV3 in kW Electrical power HP pump||++||+|
|CV4 in kW Electrical power booster pump||+||
It is desirable to control salinity and permeate flow independently of each other, i.e. the controller is supposed to decouple the two interacting loops. In order to achieve these control targets, two manipulated variables can be used:
The resulting 4x2x0 MPC structure with four controlled variables (CV), two manipulated variables (MV) and zero disturbance variables (DV) is shown in Table 1.
In each column, the effect of the respective manipulated variable to all controlled variables is marked. “++” means a strong positive effect, “-“ a weak effect with negative sign. Permeate salinity has a very high weighting (highest priority) but will stay inside its large dead zone during normal operation.
The MPC will focus on permeate flow. Electrical power values of the pumps have much smaller weightings (lowest priority). The setpoints for electrical powers are the optimal efficiency operating points of the respective pumps.
The ModPreCon function block of PCS 7 Advanced Process Library block is ideally suited for the described tasks. The MPC Configurator provides automatic MPC design, using a few transparent parameters for adjustment of the dynamic behaviour. The conventional PID controllers and the MPC automation are implemented in SIMATIC PCS 7 and connected to the simulator software.
First, in a simulation without fouling the ability of the MPC to cope with the multivariable control problem and achieve stable feedback control is checked.
A setpoint step in throughput (Figs. 2 and 3) shows good tracking performance of the MPC. The effect of prioritised control will be shown in a situation where salinity is at its limit and a load change is demanded (Figs. 4 and 5).
In this situation, the MPC moves salinity back into the control zone, tolerating some deviation in permeate flow. The simulation results are promising. Higher degree of automation, more stable process operation with smaller variations in all characteristic values, higher safety of compliance with quality specifications, i.e. salinity will stay in its range, and reduced specific energy consumption can be expected from the MPC solution.
In another requirement context, the MPC could be used to obtain higher average throughput. In this case, the value of the quantifiable benefits (energy savings) might well be exceeded by the non-quantifiable benefits.
For example, a higher degree of automation of each reverse osmosis unit in a plant might be the prerequisite for a systematic optimisation at plant level, e.g. load balancing of all units in such a way that more load is distributed to those units that are currently in a cleaner state with respect to fouling and therefore currently provide higher efficiency.
Similarly, a higher degree of automation might be exploited for adaptation of plant throughput to varying energy prices
(e.g. day and night) in conjunction with the capacity of tank farms.
 Desalination industry enjoys growth spurt as scarcity starts to bite. Global Water Intelligence. www.globalwaterintel.com/desalinationindustry-enjoys-growth-spurt-scarcity-starts-bite/
 Desalination plants worldwide, desalination technologies, costs and the environment. 2014. www.global-greenhousewarming.com/desalination.html
 Siemens AG: SIMATIC PCS 7. www.siemens.com/simatic-pcs7
 Siemens AG, Industry Sector, Industrial Automation: White Paper “How to Improve the Performance of your Plant Using the Appropriate Tools of SIMATIC PCS 7 APC-Portfolio?”, 2008. www.automation.siemens.com/w2/efiles/pcs7/support/marktstudien/WP_PCS7_APC_en.pdf
 E Field, KL Howe, BM Thomson: Effect of Solids Retention Time in Membrane Bioreactors on Reverse Osmosis Membrane Fouling. University of New Mexico, Report to New Mexico Environment Department, Apr. 2010. www.unm.edu/~howe/UNM%20Howe%20MBR-RO%20Report.pdf
Contact Keshin Govender, Siemens, Tel 011 652-2412, firstname.lastname@example.org
The post Optimisation of seawater desalination by advanced process control appeared first on EE Publishers.
Source: EE plublishers