Novel methods for monitoring wastewater from the wastewater network and across the sewage treatment plant to aid optimisation

Even with business as usual, the challenges facing a wastewater treatment plant operator are considerable. In activated sludge plant treatments how do you maintain the correct balance of microbes? How do you optimise treatment processes? What about when the influent hydraulic and organic loads vary daily, weekly and seasonally and with storms? All of this must be achieved whilst also reducing environmental damage from discharges and reducing pollution incidents, and remaining within discharge limits. Added to these challenges is the introduction of smart network control with influent sewers. All this takes place in the context of what are sometimes seen as inflexible and mechanical operating regimes. Indeed there are multiple drivers for the optimisation of treatment processes, not least high energy costs, so how can operators achieve optimisation of wastewater treatment processes?

Luckily, with the digitisation of the water industry, data from on-line instrumentation is increasingly being used for process optimisation and control and for monitoring compliance critical operations. Online instrumentation is therefore becoming fundamental for managing risk, maximising efficiency and providing the wider business with reliable and trusted information. However, monitoring data can only be trusted from instrumentation that has been appropriately selected for the application (taking into account whole life cost), properly installed, maintained and operated. Measurements in crude sewage at Sewage Treatment Works (STWs) are a particularly challenging environment. This article will outline a few examples of novel solutions for measurement in challenging wastewater applications.
In the typical wastewater and industrial effluent treatment process, monitoring for process control, optimisation and compliance takes place throughout the works from the inlet to the final effluent and at every stage of the treatment process between these two points. Monitoring also takes place in the sewerage network before the wastewater reaches the plant, for example at sewage pumping stations to ensure the pumps are operating effectively and to flag up ragging issues.  
The current UK investment cycle for the UK water industry is from 2020-2025 (AMP7) during this period the water authorities publishes its plans to meet their key objectives for the AMP. One of the drivers is to encourage water companies to adopt methods to increase the lifetime of pipes;  rather than replacing rising mains try to operate them more efficiently. New methods for monitoring rising main performance such as high sample rate pressure monitors [diagram 2] are helping in reaching these drivers.  Sewage pumping stations are fed electrically by a control starter section, sewage is pumped up a rising main for up to several kilometres into a gravity sewer where later the sewage is treated at a water recycling centre. This process is energy intensive, and there are several factors which impact the performance of the system as a whole. Issues such as the material, length and topography all pose challenges on how the rising main performs.

Anglian Water was looking for a solution for monitoring sewage pumping stations that would be cost efficient, identify rising main bursts and help to identify areas of opportunity to optimise the system. Anglian Water worked with two suppliers to develop a burst alert and then use the high sample data to understand the performance of the rising mains. Being able to collaborate like this has helped in understanding how the pipe work is performing, as well as being able to have knowledge if there is a remote failure in the system.
Anglian Water are also undertaking advanced monitoring and control of their pumping stations. They are using signals from non-contact level monitors alongside intelligent pump starters to detect blockages on pumps and issue pump reversal command to prevent blockages.
During trials by Anglian, the automated control avoided 11 failures within a 5 month period.  Anglian Water are focusing on installing this new solution at pumps at higher risk of failure. It can provide early warning of pump failures and enable planned maintenance rather than reacting and doing the maintenance after a pollution event. The controller also provided flow measurement through the means of rate of change and size of wet well. The accuracy is further increased by the monitoring of rising main back pressure and in putting the pump curve.
“The new solution developed is more appealing as it results in a lower cost solution than traditional flow measurement solutions which tend to cost more in terms of installation and distribution interruptions,” explains Lorenzo Pompa, Asset Optimisation Engineer at Anglian Water, “ there are limitations to this system because it relies on data flows from various sensors – for pressure and level, but overall we are happy with the outcome of this project and it gives us the information we need to make the appropriate intervention”.

Once wastewater actually reaches a plant (diagram 1), ammonia and dissolved oxygen are usually measured across the wastewater treatment plant to measure and control treatment processes. Compliance levels for effluent discharges are set at approximately a 70-80% reduction in Biological Oxygen Demand (BOD). It can be argued that BOD is a far more effective measure to assess whether water is ‘safe’ for discharge compared to ammonia alone, because it measures the natural oxidation and recovery conditions in the receiving water. BOD measurement is broader than measuring Disoolved Oxygen (DO) as it encompasses all oxygen consuming constituents of the wastewater. Previously, turnaround for BOD lab tests would take 5 days. However, a multi-parameter probe that uses fluorescence to monitor BOD, Chemical Oxygen Demand (COD) and Total coliforms in real-time is now available. Measurement of ‘tryptophan like’ fluorescence can be used to infer BOD using an algorithm which includes compensation for temperature and turbidity. A major UK water company has used this probe to measure BOD in real time and to control their activated sludge processes. The correlation between grab samples tested in laboratories versus online BOD measurements was very good. The trial found that aeration control based on BOD was better than using ammonia values. There is a potential cost saving for aeration compared to fixing the speed; real time BOD measurement can be used to optimise treatment to use consent limits more fully and ensure greater process stability. The technology can be used for feed back or feed forward approaches. The manufacturer claims that this system represents a true early warning system based on direct measurement rather than secondary parameters such as DO or ammonia.  It can allow users to identify potential problems long before the process or ecosystem is out of control.

Another new combination technology has been developed, this time for automated optimisation of the activated sludge process (diagram 1), and is already in use in STW in Ireland, the Far East and in Europe. An online sensor array of the usual parameters (dissolved oxygen, ammonia, pH) [see diagram 1] provides real time monitoring which is self-calibrated using automated grab samples. The system is self-cleaning by having separate core and shell components which move against each other to remove rags, and wipers run every 30 minutes to clean the sensor array. This combination technology can be installed on both the inlet and outlet, and can provide real time control of activated sludge and energy savings of up to 18% of the total site energy. In addition, the data generated from this technology can be used to detect issues occurring on the plant by interpreting changes in data. For example toxicity in the works can be detected by a decrease in the usual oxygen uptake rate (OUR) by bacteria on a plant.

The OUR will be site specific based on the biodegradability of waste present. In fact operators have found that measurement of oxygen uptake can be the most powerful aspect of this technology, to calculate the feed forward load and guarantee compliance within discharge permits.
In comparison, a radically novel, simple and cost effective approach has been developed for online monitoring of aeration lanes in the activated sludge process. Colormetric analysers, commonly used at the final effluent of waste water treatment plants and in environmental monitoring, have been adjusted to cope with the sampling difficulties found in areas of the waste water treatment process. These have been installed in STW in the US, with on-site adaptations like heat tracing lines to keep the equipment functioning at lower temperatures. Online monitoring of orthophosphate using colormetric analysers showed a good correlation with data from grab samples, as did online monitoring using an ammonia analyser (when used in a membrane bioreactor). Design features for this online wastewater monitoring system include large diameter sample lines to minimise clogging, whilst microfluidic reagent tubes are used to minimise use of reagent (typically 5-20 litres per year), use of a replaceable 50 micron filter (typically lasts 2 months and costs $25 to replace), and use of an easy to use and low maintenance vacuum pump.

Another new method to monitor microbial processes in real time makes use of a biological sensor, which is already in use in the US, Canada, China and around Europe. The biosensor reacts to changes in environmental conditions in real time to produce changes in electrical current (Microbial Electron Transfer or MET); enabling you to measure microbial activity in real time. The biosensor has recently been trialled by United Utilities to help manage an activated sludge wastewater treatment plant. MET monitoring data was found to correlate with organic loading very well, with rainfall events resulting in greatly reduced MET. MET real time data correlates well with BOD and Total Organic Carbon (TOC). MET also tracked blower use very well and could be used for optimisation of aeration. Therefore MET values could be combined with energy usage to calculate overall treatment efficiency.  The three month trial concluded that MET data provides a weekly pattern for understanding baseline loads moving through the system, and can assist in guiding the centrate loading schedule. The biosensor can also be used to identify toxic events hitting the plant. Over three months no maintenance was required. Possible next steps would be to combine MET data with flow data and Mixed Liquor Suspended Solids (MLSS) to make a weighted estimate of organic loading to assist in aeration and sludge return. MET data could be used to monitor Activate Sludge Plant (ASP) performance as a process tool, and understand which conditions are resulting in more efficient treatment. MET monitoring could be used to quantify loading from the centrate used, and possibly to define the value of receiving centrate at the plant.

This is just a quick overview of a few new technologies available for measuring challenging wastewater applications, based on technologies discussed at two recent Sensors for Water Interest Group (SWIG) events in January and March 2020 (www.swig.org.uk ). It is also worth considering the criteria for adoption of new technologies by water utilities. In the UK at least, bidding criteria for suppliers has changed with the latest AMP7 cycle of investment. Box 1 summarises the current criteria for instrumentation selection, installation and commissioning as outlined by Richard Bragg, Principal ICA Engineer at United Utilities.

“Ideally we are looking to reduce the cost of process data, also lower power and better comms, better diagnostics, lower skilled maintenance and use of the Ethernet” explains Richard Bragg, “conversely it may come as a surprise that we don’t need ‘clouds’ as we have our own and we don’t need too much new technology necessarily either.” Mr Bragg questions whether there is actually a need for more instruments, as there is now more use of modelling, digital twins, AI and machine learning to better gain insight from monitoring data from existing instruments.
The final note about measurement in such a challenging environment goes to Richard Bragg again “I would ask any supplier of a new instrument to prove that it will actually work in wastewater applications, and that maintenance models are realistic”.