I have an outstanding piece of software, called GPS-X that I use for modeling just about any type of wastewater system. As an example of the type of modeling you can do, and the level of detail you can create with your model, look at a process flow diagram of the “typical” unit processes that might comprise an industrial wastewater treatment system shown in Figure 1. I’m showing you this graphic because I want you to get a sense of the tremendous capabilities of this very sophisticated wastewater modeling program. We’ll pick a simpler model and do some modeling so you can see the output you can generate.
Figure 1: Complex Wastewater Model
An oxidation ditch wastewater system is shown in Figure 2. The MLSS goes to a splitter box where it gets distributed to two secondary clarifiers. Waste activated sludge goes to two aerobic digesters operated in series. Digested sludge is pumped to sludge drying beds. The model created in Figure 2 has been built using the actual dimensions and volumes of the unit processes shown. So we are modeling an actual system using real design parameters, flow rates, and loading data.
There are several treatment issues this particular wastewater plant has to deal with which include pH, temperature, ammonia, and organic loading. The biggest issue is the organic loading and that’s what we’ll focus our modeling effort on. Directly associated with the problem of high organic loading is oxygen depletion in the oxidation ditch. So we are also going to focus our modeling on oxygen consumption rates.
Figure 2: Oxidation Ditch Wastewater Model
Once you’ve built your model, which is what we did in Figure 2, you are ready to run simulations, during which you can adjust any parameter you want such as increasing or decreasing the influent flow rate, COD concentration, ammonia concentration, air flow into the oxidation, etc. The four panels of the simulation screen for GPS-X is shown in Figure 3. There’s a lot of information crammed into this graphic so if you click on the graphic a larger view will open up for you. I’ll give you a closer view of the input screen which you can see in the upper left corner and I’ll show you some of the outputs that are generated on the two panels on the right side of Figure 3.
Figure 3: GPS-X Simulation Screen
Figure 4 shows the inputs that I’ve selected for running my simulations. Keep in mind, I picked these particular inputs because the influent COD, ammonia, and flow are the key variables that I need to model for this wastewater system. I could have chosen any number of other (or additional) inputs. Look at the COD input in Figure 4. The starting point shows an influent COD concentration of 500 mg/L. There is a slider to the right that allows me to increase or decrease the COD concentration while I let the model run. With the values you see in Figure 4 selected, let’s take a look at some output before I run the model again and start changing the input values.
Figure 4: GPS-X Inputs
A close look at Figure 5 shows you several important parameters and their values as predicted by the model. We can see that the MLSS, given the flow, loading, and sludge wasting rates selected in Figure 4, will stabilize at around 2,661 mg/L and this value is not too far off from how the oxidation ditch actually operates. But our DO is low at just 0.086 mg/L and this low value also matches very well with actual operating conditions in the oxidation ditch.
Figure 5: GPS-X Simulation Result Based on Inputs Shown in Figure 4
If I want to improve the dissolved oxygen concentration in the oxidation ditch I have several options that I can model. I can reduce the COD loading to the biological reactor; I can increase the volume of the reactor; I can decrease the flow to the reactor; or I can increase the oxygen input by increasing the horsepower and/or the efficiency of the aeration equipment in terms of its oxygen transfer rate.
Now let’s be realistic. I don’t know of too many wastewater plants that can reduce their COD loading, at least not in the long term. Industrial wastewater plants often do have various “off-spec” tanks they can divert flow to for a short period of time. But at some point that flow is going to have to be reintroduced to the wastewater plant. Nor can plants reduce their flow. So really, we should consider the flow and COD concentration and loading to be fixed. In terms of time and money the modeling is very useful in pointing out the need for major capital improvements such as adding aeration capacity (volume). And the modeling has been used very successfully to do just that. But in the short-term the best option is going to be to add oxygen input capacity. This can be done by adding floating aerators, skid-mounted diffused air systems, and even hydrogen peroxide.
So let’s model the oxygen input generation capability and see just how much aeration horsepower we need. I’m doing this in terms of horsepower because this oxidation ditch uses a combination of rotating disk aerators (Figures 6 & 7) and “boat aerators” (Figures 8 & 9).
Figure 6: Oxidation Ditch Rotating Disk Aerator Off
Figure 7: Oxidation Ditch Rotating Disk Aerator On
Figure 8: Example of a Boat Aerator
Figure 9: Another Example of a Boat Aerator
In Figure 10 you can see the result of my having increased the aeration horsepower input to the oxidation ditch. I wasn’t willing to keep it simple though so at the same time I also reduced the oxygen transfer rate because I find manufacturers of aeration equipment tend to overestimate the efficiency of their equipment. This is just my opinion and it is not based on any scientific, field-verified testing. I just want to be conservative in my modeling and the estimates I generate.
So what, exactly, did I do in the model. Well, GPS-X breaks an oxidation ditch into 16 sections. In Figure 5 you can see where it says “Mech. Aeration Power hp” I had 20 hp listed giving a total installed horsepower of 16 sections × 20 hp/section = 320 hp. In Figure 10 you can see that the hp in each of the 16 sections has been increased to 30 giving a total installed horsepower of 16 sections × 30 hp/section = 480 hp. But I dropped the oxygen transfer rate from 5.76 lb O2/hp-hr to 4.94 lb O2/hp-hr. Neither of these values are unreasonable as you can tell by looking at the clean water transfer rates shown in Figure 11. Still, you can see that our DO concentration went from an unmeasurable 0.09 mg/L to 0.94 mg/L which represents a 10-fold increase in the oxygen concentration. The point is, in making such changes to different variables, you need to go through many iterations. And though it may sound like a lot of work, it is not. There is no more efficient way to generate so much data and insight into how best to modify a wastewater system than through a powerful model of that system. And GPS-X has all the tools and all the power you could possibly hope for.
Before concluding our discussion on the GPS-X wastewater modeling program there’s just one more graph I want to discuss and it’s shown below in Figure 12. This graph was created so I could monitor changes in the dissolved oxygen concentration in the oxidation ditch while I made adjustments to the COD concentration while running the simulation. You can create all kinds of graphs like this to watch how a given parameter or variable will change as a function of changes being made to another parameter. Obviously, we know that in a biological treatment system the organic load is going to have a very direct impact on the oxygen concentration in the reactor as this graph clearly shows.
Figure 10: GPS-X Simulation Result After Increasing Aeration Horsepower
Figure 11: Oxygen Transfer Rates - Clean Water
Figure 12: DO Concentration vs. COD Loading