Michael Dooley explores resilience, big data, AI and optimisation of the AS process
The activated sludge (AS) process is one of the most important environmental protection technologies in the world. Some 77,000 AS plants are in operation in Europe alone, protecting some of our most critical receiving waters. The process has a significant carbon footprint, consuming over 1% of the power generated in developed countries.
The published definition of ‘resilience’ by OFWAT is “the ability to cope with, and recover from disruption, and anticipate trends and variability in order to maintain services for people and protect the natural environment now and in the future.” From a control and instrumentation point of view, the development of a truly resilient AS process requires:
* Online monitoring of influent loads.
* Control of the treatment process to treat this load with the lowest possible carbon footprint.
* Online monitoring of discharge load to protect the receiving environment.
* Effective management of data and operational trends to investigate and determine the root cause of operational issues, feeding back into continuous improvement of Asset control and maintenance.
However, the provision of effective instrumentation in the AS process has long been a challenge with calibration drift, sensor head fouling and long-term reliability offering specific application challenges.
The use of oxygen uptake rate is a hugely effective measure of both influent and effluent load, and when combined with ammonia, as an online monitor, can be used to give effective feed-forward /feed-back control of the process.
Avoidance of calibration drift is a design requirement for the sensors themselves but ultimately, reliable operation will depend on the effectiveness of cleaning, maintenance and calibration regimes at the end user premises. During a recent conference in the UK it was presented that simply cleaning and calibrating DO sensors in the AS plant would lead to a 10% reduction in aeration energy consumption.
The exampleXXSHOWN BELOW/LEFT/WHEREVERxxx is from a fully automated high point calibration of the dissolved oxygen sensor in the Strathkelvin ASP-Con system. Prior to calibration the sensor is automatically cleaned. As the process is automated it can be timed to occur when loads on the plant are low and can also be performed at a much higher frequency than a manual calibration.
The ASP-Con system has fully developed automatic high and low point calibration for dissolved oxygen, pH, ammonium, nitrate and ORP. When combined with automatic cleaning of the sensor heads every 12 hours, we have an effective basis for tight control of the treatment plant set-points.
The output trend xXshown below/left/whereverxx from an ASP-Con installation in 2014 indicates over-aeration of the process when ammonia loads are low, giving opportunities for aeration energy savings. Only when plant operators have confidence that these readings are accurate and repeatable can they implement full optimisation regimes.
Now that we have effective feed-forward and feed-back load measurement, reliable and accurate measurement of the critical set-points and online monitoring of bacterial measures of performance such as settlement, MLSS and SVI we can turn our attention to the integration of the instrumentation into the plant PLC-based control systems.
Connectivity of instrumentation has been considerably simplified using MODBUS/PROFIBUS protocols so the remaining challenge turns to data management and the control strategies, which can be applied. This is undoubtedly the next stage in the development of truly resilient wastewater assets.
Strathkelvin Instruments is co-operating with AEMA in Spain to develop an advanced real-time control strategy based on the ASME3 model for management of the AS process due for release in Q1 2018. Current control strategies are being adapted on several of the company’s client sites to introduce decision making, variable set-points and aeration energy optimisation based on the increasingly detailed and accurate information available from the advances in instrumentation provided by the ASP-Con system.
One question remains: how far can this change go? Does it eventually lead towards the adoption of artificial intelligence (AI) for the AS process? This is a difficult thought process! On one hand, the process being biological in nature is very difficult to model under all circumstances and there is a requirement to zealously manage risk to the receiving environment.
On the other hand, the risk to human life is not immediate from failure of an AI control system (such as for example in AI automated heavy goods vehicles) and the threat of significant job losses is unlikely (although we do expect an upskilling of operator tasks and capabilities).
The benefits of fully tapping into advanced real-time control of the activated sludge process are improvements in discharge compliance from early detection and management of upset conditions, reduction in carbon footprint due to the reduction of control system margins and direct cost savings due to energy consumption reduction.
Applying the techniques described above and combining them with resource and energy recovery in the process could lead to wastewater treatment regularly being a net energy generator. Due to the regulatory and costs pressures, the industry is slow to take up new technologies so it is likely to be many years before the question is fully resolved. In the mean time there are both huge global environmental and costs benefits to be gained from fully adopting the improvements currently available.
Michael Dooley is managing director of Strathkelvin Instruments.