Modelling Problems
To summarise the sections “Case Study Site: WWTW capacity” and “Wastewater Time Energy Profile” it can be seen that load shifting to allow further renewable integration may be possible by storing wastewater for processing at a desired time. For the WWTW case study site a capacity in terms of energy consumption of 54kWh per hour was deduced with the plant operating below this 97.2% of the time with an average hourly energy demand of 39.7kWh. Note the plant has been recorded as operating at up to 75.4kWh but this was taken to be undesirable. It was then assumed the plant has a minimum constant baseline hourly energy demand of 15.76kWh with the remaining energy consumption representing the volume of wastewater being treated, with each metre cubed of wastewater requiring 0.4254kWh of energy to treat. This 0.4254kWh of energy was then taken to be required across 14 hours and distributed as shown in Fig.8 and Table 4.
Assuming the wastewater at the site simply flows into the WWTW, is treated in the given 14 hours, and flows out in a continual process the next step in attempting to model this method would be to convert the hourly energy demand for the year, shown in Fig.5, to a wastewater flowrate. However, at this point the data does not agree with what would be expected.
Domestic water flow and demand generally follows a diurnal pattern with a peak in the morning and the evening and lower values overnight. A typical flow pattern showing domestic water demand is plotted in Fig.9 with the typical expected percentage of a day’s total water demand for each hour shown. (16) Overlaying the average energy consumption for WWTW case study site from Fig.2 it can be seen that the energy demand roughly follows a similar pattern with the WWTW demand leading by around an hour.
Percentage of daily water demand for each hour of a typical day (16) with data from Fig.2 showing the average hourly energy demand of WWTW case study site overlaid.
There are two points to note, however. The first is that based on the assumptions made so far, a slightly different pattern would be expect. It would be expected follow a similar pattern but shifted further along the time axis as the aeration process does not occur until several hours into the treatment process. Therefore, the spike in energy demand would be expected to also occur several hours after the morning peak in demand. Although the site’s energy consumption is shifted, it is only by an hour and is 10 significantly less than would be expected. Given the delay between demand by customers and the water used arriving at the WWTW for processing the energy demand at the case study site appears to almost vary according to the flowrate into the plant. Although there will be an increase in energy demand as water is pre-treated and pumped upon arrival at the site the peak in energy demand would still be expected to coincide with the aeration process which has been taken to occur following the six-hour primary clarification. Fig.10 shows an estimate of the expected average energy consumption at the case study site, with a shorter 3-hour clarification process also shown to consider the effect of the length of this process. This is based on the water flow demand shown in Fig.9, (16) assuming the case study WWTW treats on average 1.35m litres per day, and using the energy profile given in Fig.8.
Expected average hourly energy demand for WWTW case study site.
The expected energy demand shown in Fig.10 can be compared to the case study in California. (2) Fig.11, taken directly from this case study agrees with the expected energy consumption pattern shown in Fig.10 rather than the real data pattern shown in Fig.2. Although they do not line up exactly the maximum and minimum values are around roughly the same time, give or take an hour which can be attributed to variations in processes and timings. Unfortunately, the reason behind the pattern found in the WWTW case study is that the data is unknown.
Energy demand for Laguna WWTW, California
The second point to note is that although a pattern in energy demand can be seen in Fig.3 if the axis of this plot is changed, as shown in Fig.12, it becomes clear that the energy demand is actually generally fairly consistent throughout the day and night. Again, this isn’t what would be expected if wastewater simply enters the plant, is processed, and leaves, which has been assumed so far. At some point a period of reduced energy consumption would be expected due to the lower flows associated with the night hours. Fig 10 shows the rough variation that would be expected and although Fig.11 shows a variation less than that in Fig.10 it is significantly greater than can be seen from the case study data. The difference in variation between Fig.10 and Fig.11 may be due to the flow equalisation basin used at the California site that dampens the variation in flow rate through the plant as discussed in the Discussion section.
WWTW case study site’s average hourly energy demand for January to March, April to June, July to September and October to December, representing the four seasons of the year.
In an attempt to model this concept, without access to flow data it was assumed energy consumed by the WWTW was representative of the flow of wastewater through the plant. This was the core assumption around which this model was to be built. However, if the energy consumption varies little across 24 hours, using this assumption it would be concluded that the flowrate of the case study site would be relatively consistent, but this is not what would be expected with a typical variation shown in Fig.13 taken from the Environmental Protection Agency (EPA) Wastewater Treatment Manual. (5)
Diurnal and seasonal flow variations to a treatment plant
Given the information shown in Fig.13 it could be possible that Scottish Water are already using flow equalisation basins to manage the daily flow variation of wastewater throughout the day. This will be discussed in more detail in the discussion section, but whatever the case without understanding the specifics of this site and the reasons behind the flow pattern and lack of daily variation the fundamental assumption that was being used to model this concept does not hold. Despite being unable to model this concept, it will still be discussed within our Discussion section.