By definition, waste water treatment works are extremely high energy users and their efficient running requires a fine balance of biological and hydraulic parameters throughout the process. However, maintaining that balance can be a daunting task; especially if operational management do not have access to ongoing reliable and meaningful data on essential interface levels at critical stages of the treatment processes, including primary clarifiers, secondary clarifiers and thickeners.
Hycontrol’s Nigel Allen investigates the complex process control and automation problems associated with sludge blanket level and interface monitoring within waste water treatment and allied industries. Allen further outlines how the latest sonar technologies overcome traditional interface monitoring shortcomings by simultaneously monitoring RAS and FLOC levels to optimise plant control and bring significant energy savings.
Accurate measurement of interface levels is a complex problem in the murky, turbid settling tank environment and without extensive sample extraction and subsequent lab analysis it can be extremely difficult to obtain a clear picture of the all-important density profiles. The sludge within the tank decreases in density as you move from the bottom of the tank towards the top water level. The densest sludge, sitting at the bottom of the tank, can range from 3000 to over 6000 mg/L and in a stable tank the sludge will gradually decrease in density to around 200 mg/L at the top of the column. Generally, treatment works are interested in ‘quality’ sludge which has a density greater than 2,500 mg/L. This sludge at the bottom of the tank is referred to as RAS (Returned Activated Sludge). Sludge at this density is heavy enough not to move up the tank when hydraulic or biological problems occur and is also dense enough to be termed “good quality” biomass, which can be returned to the aeration lanes to help with the pre-treatment process or diverted to waste. However, when a change occurs to the site loading process problems can occur and operators need to know the dynamics of the different interfaces to assess and effectively control the ongoing process.
Many sonar systems struggle to provide comprehensive and reliable information under these difficult conditions because they do not have the power and the correct frequency to penetrate through the suspended solids. In the absence of anything better, the only other way of gaining a full ‘top to bottom’ picture has been to use manual dipping products such as a ‘Sludge Judge’ or a gap sensor.
However these labour intensive devices do not provide a continuous output for trending and control. They only give a visual snapshot of the interface layers in the tanks, whilst having associated and undesirable health and safety issues.
As outlined above, traditional sonar interface monitoring systems fall well short of the necessary requirements. Their frequency range and lack of power means that they cannot penetrate much further than densities of approximately 1200-1500 mg/L, thereby only allowing them to identify the upper FLOC interface with any level of certainty. Based on this information, a huge and potentially catastrophic assumption is then made that the corresponding denser RAS interface tracks the FLOC interface under all conditions. Under stable conditions this is indeed what happens (see Fig. 2 & Fig. 3). However, when imbalances develop due to changes in site loading, continuing with this assumption makes matters even worse.
If the site is using an instrument that can only monitor the lighter density FLOC layer as a basis to control the RAS pumps then, when the FLOC layer rises due to an imbalance, operators are automatically assuming the denser RAS layer is also rising. As a result the site will inevitably either increase the RAS pumping rate or drop the bellmouth in an attempt to bring the rising blanket back down the tank.
However, what is actually happening is that the denser ‘good quality’ biomass has remained at the bottom of the tank and it is only the lighter FLOC layer which has lifted.
Increasing the pumping rate or dropping the bell mouth will have very little affect on the lighter FLOC layer, which has risen up the tank and these actions will very quickly remove all of the ‘good quality’ biomass from the tank and then begin to pump back a lighter density ‘poor quality’ biomass. This will subsequently increase the problem by having a negative effect on the F:M ratio (Food to Micro-organisms Ratio*).
*The F:M Ratio is one of the fundamental control parameters for the activated sludge process. The ‘food’ in the ratio is the CBOD (Carbonaceous Biochemical Oxygen Demand or Carbonaceous Biological Oxygen Demand) entering the process, the ‘micro-organisms’ are the activated sludge solids in the aeration tanks, which are measured as ppm or mg/L of MLSS (mixed liquor suspended solids). To establish and maintain a consistent CBOD and secondary waste removal from raw sewage, an activated sludge process must maintain the weight of food to weight of microorganisms under aeration.
On some sites it could take weeks to fully rectify the situation and during this time increased aeration may have been required increasing energy consumption at the plant and therefore energy costs. To optimise plant efficiency it is essential to monitor the ‘good quality’ biomass at 3,000 to 6,000 mg/L and the FLOC level simultaneously. This allows the possibility of automatic control of RAS pumps and bellmouth valves to ensure that ONLY ‘good quality’ biomass is returned back to aeration or to the thickener for wasting.
There is now a highly effective sonar system which ensures such situations cannot occur, by simultaneously monitoring both interfaces. Sonarflex’s submerged high power transducer sends ultrasonic pulses through the liquid, which are then reflected back from the different density interfaces and are even powerful enough to penetrate densities in excess of 6000 mg/L and detect the tank floor. These signals are processed by the specialist software to provide outputs relating to both the FLOC and RAS levels within the tank. This vital information forms the basis of improved process and control to enable the site to optimise energy consumption and site operations. Alarm levels can be set so that in the event of the FLOC level lifting, operators can make the necessary process changes in plenty of time to prevent the problem continuing and avert a breach of consent.
The key to the success of this innovation is the availability of a wide range of transducers, with frequencies ranging from 30 kHz to 700 kHz. Comparing the theory for ‘through air’ ultrasonics, (which is a well established technology for level measurement) it is possible to understand the need for multiple frequencies in sonar applications. Measuring the level of a simple liquid in a vessel 10 metres deep is very straightforward and almost any high frequency (40-50 kHz) transducer will give reliable and repeatable results. However, if we use this same frequency on a similar size silo that contains a solid such as cement, with high airborne dust concentrations, then the results are far from successful. It will inevitably struggle to penetrate more than a few metres and would be highly unstable in fill conditions because the suspended particles will attenuate the high frequency short wavelength signal.
By comparison, if a lower frequency (5-10 kHz) is used with a longer wavelength, then the sound wave can pass through the suspended particles more easily. A perfect example of this is the use of a foghorn. In bad weather conditions visibility is poor because the air is saturated with moisture. A high frequency – short wavelength would be far less effective in this scenario as the sound would be attenuated by the moisture particles and only travel a short distance. Foghorns use a low frequency – long wavelength to project the sound through the moisture particles miles out to sea to warn ships. This is known as the ‘Foghorn Principle’.
This same analogy remains true for sonar. Whilst traditional designs adopt a ‘one size fits all’ philosophy for sludge blanket systems (adopting a range around 600-700khz), the optimum transducer frequency needs to be selected to ensure the best engineered solution across a treatment works. Sonarflex uses a different frequency transducer for primary sedimentation, primary and secondary clarifiers, sludge thickeners, lamella clarifiers and sequential batch reactors (SBRs).
SBRs are typically installed where space or cost are at a premium. They combine the primary sedimentation tank, the aeration process and final/secondary settlement all in one tank. By the nature of the principle of their operation, the liquid levels change within the tanks and a traditional fixed transducer cannot cater for these changes. To overcome this, a unique floating transducer is used enabling it to track the settling blanket interface as decant levels change. As a result, settling times can be monitored far more accurately and the improved batch times can increase throughput by up to 20%.
The waste water treatment environment is harsh and standard instrumentation designs stand little chance of surviving more than a few weeks. The ultrasonic transducers (either submerged or floating) need regular cleaning to avoid unreliable performance due to signal attenuation caused by the build up of scum, scaling, air bubbles or fats. However mechanical cleaning systems, such as wipers, have a finite life and require constant maintenance and components can often need changing as often as every few weeks. To overcome this, the Sonarflex uses a patented actuator lever arm system. (Cleaning image somewhere here) The automatic cleaning cycle is triggered on a time basis or by a predetermined reduction in signal level. When this occurs the actuator pushes the transducer support arm through the water to an angle of 45° and then returns it to the vertical. This sharp shearing action through the water removes any debris or scum from the front face and ensures optimum performance, without the need for any operator involvement.
The attention to detail in the design of the Sonarflex covers both electronic and mechanical operational features. There is a hazardous area ATEX version, which can be used for the growing number of enclosed settling tanks, built to minimise odour release to the atmosphere or to capture and reuse the methane gases. A wide range of communication protocols including Fieldbus, Profibus, HART and DeviceNet, ensure seamless integration with modern plant instrumentation and DCSs. The transducer can be located up to 500m from the control unit and a robust wireless link option can provide communication for rotating bridge fitted units, whilst GSM connectivity provides instant access to all parameters for servicing, technical support and commissioning. Multiple outputs and relays can be used for alarm and control functions as well as cleaner arm actuation.
By providing both analogue and wide range of BUS communication protocol options as outputs Sonarflex can help to maximise the efficiency of the process. When using the analogue version of the instrument, two 4-20 mA outputs are available for monitoring the different densities within the tank. On a primary tank the interface can be monitored using one output, whilst suspended solids between the transducer face and the interface can be monitored using the CLARITY output, providing an indication of how well the tank is settling. This second output can be used to optimise dosing control by dosing only when needed and not on a traditional timed basis, the amount of flocculent or coagulant used can be reduced and important cost savings made.
On a secondary tank (Secondary tank image somewhere here) the two outputs can be used to monitor the RAS layer and the FLOC layer providing control of the RAS pumps or bellmouth to optimise the density being returned to aeration and ensure a consistent density is wasted to the thickener. The lighter density FLOC output can provide indication of process problems and also provide an early warning of a possible breach of consent. In a thickener the two outputs can be used to monitor BED level and water CLARITY. Monitoring the BED level ensures the filter presses or the digester receives sludge of consistent density with low water content, from the underflow pumps. This reduces foaming and mechanical wear and tear, therefore making the process more efficient. Monitoring CLARITY (suspended solids between the transducer face and the BED level) provides a control for dosing whereby the instrument will provide an output indicating the concentration of suspended solids. As suspended solids increase, dosing can be increased and as they reduce, the dosing can also be reduced, maximising the dosing process and reducing waste from over dosing.
Alternatively, if the instrument is utilised using the comms options then the plant PLC can receive three outputs with any combination of RAS, FLOC, BED (level), and CLARITY being available. Whether operating the instrument using the two analogue outputs or the four comms outputs, the valuable information provided can improve control through the works, provide rapid indication of process problems, prevent breaches of discharge consent, control dosing in primary tanks and thickeners as well as reducing wear and tear on filter presses. All of the above can help overall to reduce energy consumption, maintenance and chemical costs on sites.
As Allen concludes: “Until now operators have been literally ‘operating in the dark’ when it comes to having reliable data on critical interface levels in a range of tanks at various parts of their process. Our solution now gives them data they can use with absolute confidence for the betterment of their process. Energy savings at plants where Sonarflex has been installed around the world are very impressive and we expect UK companies to see similar advantages within a relatively short period.”
Can be contacted on