The importance of flow and load surveys
Although seemingly expensive, a flow and load survey can save vast sums of money - and it can prove to be a key aspect of a wastewater treatment works, writes Paul Lavender of Aqua Enviro
A flow and load survey should be an essential requirement of any significant wastewater treatment plant new build scheme. It is also extremely helpful at existing sites either where the works are failing and a remedial solution is required, or where uprating is planned, for instance to meet tighter discharge consents.
Although ostensibly capital intensive, this expenditure is small in comparison to the potential savings that well planned flow and load surveys can deliver. However the key is in the planning and both the equipment employed for the survey and the analysis undertaken, must reflect the project aims. This in turn requires that these aims are well defined and conveyed clearly to the contractors undertaking the survey.
The basics of planning a flow and load survey seem simple, but the reality is different and many factors can lead to poor data. These include difficult monitoring/sampling points and failure to consider all the streams that need to be accounted for (storm overflows, return liquors and so on). It is therefore essential to carry out an initial site survey to confirm the suitability of monitoring and sampling locations, to consider other operational factors that could impact on the survey and to assess health and safety issues.
The two greatest costs associated with a survey are staffing and sample analysis. But, in the absence of reliable flow data, the survey is of little value and may require extending or repeating. Needless to say, flow monitoring is one of the hardest parts of a flow and load survey and so an approach that is right first time must be employed.
Flow monitors routinely suffer from ragging/silting as well as more prosaic issues of technical problems or battery failure. For this reason it is recommended that studies do not rely on a single flow monitor, as the cost of a second flow monitor is negligible compared with the costs associated with incomplete or inaccurate flow data.
The most commonly used monitor type is an area/velocity monitor that uses a probe mounted in either a pipe or open rectangular channel. This records both the water depth and the velocity. Flow is calculated from the depth and velocity by programming in the pipe or channel size. Monitors typically take readings at two-minute intervals. These loggers can be susceptible to both silting and ragging so regular depth and velocity calibration is required.
The flow monitor uses a Doppler signal to record the average velocity from a cross section of the flow, so a number of handheld velocity readings are taken across the channel/pipe as the velocity will be greater in the centre and slower towards the edge.
Where a calibrated weir or flume is present, then an ultrasonic depth flow monitor can be used. This is mounted on the wall of the channel and uses an ultrasonic pulse to measure the water height before it passes through the obstruction. The depth of the water in the channel is calibrated to the height of the monitor from the surface of the water. Exact dimensions of the weir are required to allow the flow to be calculated.
If it is not possible to access an open pipe or channel a clamp-on ultrasonic flow meter can be used. These require a full pipe and sufficient section of straight pipe to avoid excessively turbulent flow. The effectiveness depends on the thickness, material and condition of the pipe. Monitoring of older pipes may be particularly difficult if internal fouling has occurred as this can block the signal.
As a general rule, the smaller the works the harder it is to accurately monitor the flow because:
- Screens are often poorly maintained increasing the problems of ragging
- Overnight flows may drop below the minimum detectable depth or velocity
- Low flows can increase deposition of silt in the channels
- A smaller budget allocated to the project may mean only a single flow monitor is used, although this is not advisable
- Less likely to be calibrated flumes and more hard to access underground pipework
A rain gauge with a data logger is essential to correlate rainfall events to flow variations. A tipping bucket rain gauge is usually used with a minimum detection of 0.2mm of rainfall. This is located on a level surface so that it is free of any possible obstructions within a 45° line from the ground. The data will show the lag time for the flow to increase following storm events. This is likely to vary significantly depending on the size, topography and network characteristics of the catchment area.
Composite samples are nearly always used to characterise the influent and are collected by autosamplers that are either time based or flow proportional, with the latter providing a more accurate indication of the load to the plant. Spot samples may be taken to supplement the composite data or these can be taken from specific process stages to assess plant performance i.e. samples of primary and secondary sludge, mixed liquor, thickened sludge, digester sludge and so on.
The sampling regime will depend on the detail required and the nature of the site. For a survey at a small, rural, domestic plant, a single daily 24-hour composite sample of the influent may suffice. At a larger works a more intensive regime is may be required. Unless detailed design information is required, surveys tend to take daily 24-hour composite samples with some days of more intensive sampling to provide a diurnal profile of the flow and load. It is important that flow data is matched to the sampling times to allow load to be accurately calculated.
Autosamplers are also prone to ragging and therefore influent samplers should be located after the screens and grit removal when possible. Consideration should be given to factors that affect how representative samples will be.
Fats, oils and greases are generally present on the surface of the water and may not be picked up by the auto-sampler. Thus for accurate FOG determinations samples should be taken from a point of turbulent mixing.
Sedimentation of particles occurs in slower moving channels and so composite samples may have lower solids (especially grit) than expected.
Rapid changes in composition, in particular where there are intermittent trade discharges, can occur in small catchments. These may be missed by auto-sampling due to dilution. Additional spot sampling or diurnal sampling will be required to accurately assess this.
Saline intrusion can also be missed where it is tidal and so the peak salinity to the works may be a magnitude higher than the average. Where salinity is important, a logging probe should be installed to supplement the data.
Contamination of final effluent samples must be avoided and so auto-samplers should be carefully positioned to avoid picking up any materials from the sides or edges of channels/chambers.
As with the sampling, the analysis regime should be tailored to suit the study. Often, the same analysis schedule is requested at each sample point, without thinking what data is actually necessary and how the data is going to be used. This can greatly increase the cost of a study. Although each site is different, the following comments may help assess the analysis regime required.
TSS, COD and BOD are used in assessing the load and are crucial in the design of any works, thus they are carried out on all samples.
Soluble COD and BOD – Illustrates the proportion of the organic load in solution and usually carried out on sets of diurnal influent samples to aid in process design and modelling.
pH – Unlikely to vary significantly in most domestic works but is so cheap it is often carried out on all samples. Fluctuations can be monitored using a hand held pH meter or by installing a logging pH meter.
FOG’s – FOG’s are of increasing concern but analysis is expensive and thus only daily composite samples of the influent and sometimes primary treated effluent are analysed to indicate if more intensive sampling is needed.
Alkalinity – Usually only required if the works is to nitrify and a daily composite analysis of the influent is sufficient.
Ammonia – Usually included in the main analysis suite along with TSS, COD and BOD and is of importance where nitrification is required or where nutrient deficiencies may exist due to trade inputs.
Phosphate / Phosphorus – Where there is little trade input and no P consent then this is not analysed.
Nitrate – Nearly always low in the influent and is usually only analysed after a nitrifying process stage to help provide a nitrogen mass balance.
Total Kjeldahl Nitrogen – The TKN to ammonia ratio does not alter greatly and daily composite samples of the influent are adequate, particularly as this is also a costly test.
Metals / specific organic compounds – Where there are traders within the catchment with consents for metals or other compounds such as pesticides, which may adversely affect the treatment process or risk breaching the discharge consent then additional analysis may be required. This would be entirely site specific and based upon the compounds involved.
Determining the project aims and ensuring the correct choices and locations of equipment are essential for a successful flow and load survey. The study gives the opportunity to investigate the response of individual process stages to changes in hydraulic and organic loading, therefore additional investigative work such as microscopic analysis, DO surveys, respirometry and settlement testing can complement the data and be valuable in future optimisation and/or process design.