Running your plant at low flows
So you may have seen the flyer about how to feed your plant at low flow. (If you haven’t, here are the links “How to Deal with Reduced Inflows into Your STP” & “How to Operate Your WTP and Disinfection Systems at Reduced Flows”)
Well, feeding isn’t the end of it. To keep things stable, you also need to set your plant up to run slowly. Proper setup for low flow operation will allow you to keep the process stable and producing consistent quality water.
The info below should help you work out how to set up and run your plant at lower flow rates than normal flows.
Low flow, what does it even do?
So my flows are low, that’s good, right? I mean the plant is designed for 100kL/day and I’m running at 10kL/day, I should be using 10x less chemical and it should be 10x better, right?
Well no… That makes sense conceptually, but it’s not actually the case.
Oh, really? That’s not particularly intuitive, is it?
Maybe not at first, but it does make sense once you know what’s going on, and there are two major reasons as to why sewage treatment plants don’t typically cope too well with very low loads.
The first is due to the fact that the majority of treatment is handled by a biomass. Like any living thing, it needs a minimum loading (i.e. food) to run stably. If you don’t provide your biomass with the minimum amount of loading, it will starve. This will cause part of your biomass to die off, some of it will take action to protect itself, and you can even cause bad bacteria to start to grow, which causes further issues.
The second reason is due to the physical limitations of the equipment used in the process. Typically you design a treatment process for its intended capacity, and most of the equipment you’ll use won’t have a massive turn-down ratio. This results in issues with flow, mixing and aeration energy, which also interferes with the biomass.
Huh… yeah ok, I think that makes sense.
Great. Let’s go over it in a bit more detail then.
Biological Process Issues
First up, let’s discuss the biological issues.
Whilst different processes have different tolerances for flow variation, they all operate using a bacterial biomass, be it via the bio-slime formed on a Trickling Filter or a Rotating Biological Contactor, or the mixed liquor suspended solids (MLSS) in an Extended Aeration Activated Sludge process, modern sewage treatment techniques rely upon biological (i.e. bacterial) removal of pollutants.
To keep things shorter, I’m going to keep the discussion mainly focused suspended growth processes, but most of the advice can be applied to attached growth processes too.
The activated sludge, the ‘biomass’, is a complex system of bacteria. It’s a living entity that requires a minimum amount of food to survive.
STP’s run at a state of near-starvation by design; that is how you get high rates of pollutant removal; the bacteria gobble up anything they can get at as soon as it gets near. This allows us to reach really low pollutant concentrations in our effluent.
This is what the food to micro-organism (F:M) ratio (a measure most operators are familiar with) represents. For an Extended Aeration Activated Sludge (EA-AS) treatment system, the ratio is typically very low, around 0.05 kgBOD/kgMLSS, which represents around 0.05kg of food for every kg of bacteria present.
Ahh yeah, I know the F:M Ratio.
So, while we’re starving the process to our benefit, you do need to keep up a minimum food loading or your biomass will die back.
In an Activated Sludge STP biomass die-off will typically result in a reduction in the MLSS concentration, assuming you’re running a constant sludge wasting rate. If you can vary your WAS (or your process wastes automatically based on MLSS) you’ll find your sludge age lengthening (and your wasting rates reducing) to maintain more biomass.
At sufficiently long sludge ages, the biomass will essentially eat itself. At that point, you don’t have an STP; you have an aerobic sludge digester.
Hang on, Lower sludge production? That sounds nice. We spend quite a lot of money on getting rid of waste sludge!
Longer sludge ages typically result in a reduction of the active biomass. This can typically be estimated by doing a Volatile suspended solids (MLVSS) test, which you typically won’t bother with under normal conditions as it’s easy enough to estimate based on the MLSS.
So while the lower sludge production does mean less sludge to get rid of, the reduction in the active portion of the biomass makes the process far more sensitive to shocks, be they toxic or rapid changes in loading, as there is no longer sufficient biomass present to accommodate them. Eventually, you may just straight up fail to comply with your license parameters, as the biological nutrient removal capacity is all but gone.
Long sludge ages also favour filamentous bacteria, which can have significant effects on your sludge settlement, causing issues with clarification processes.
So yeah, there’s a reason you typically run 25-30 days sludge age in an activated sludge system unless you really need to go longer.
Physical Process Issues
So, that’s biological. You mentioned there can be physical problems caused by our equipment?
Yeah, just to make things harder.
As I mentioned, your STP is typically designed to work best at close to its design capacity and is usually required to perform at 2-3 times this capacity by the regulator.
So you’re kind of limited by how much “turn-down” capacity you have in your plant, and unless you have a process specifically designed to turn-down, you may run into problems getting much below about 30% of your design capacity.
Some of your equipment will probably be more affected by this than other parts, and we’ll have a bit of a go through it below.
Yeah, good, some specifics are nice.
Let’s start with Aeration, which is pretty important for an extended aeration process, it is in the name after all.
Aeration systems are often one of the harder parts to turn-down unless they’re designed to do so, making them especially difficult to manage at low loading rates.
Generally, your aerator will be designed to supply a certain amount of oxygen as efficiently as possible, as it’s one of the biggest power users in your plant. This makes them efficient at their design point, and often far less efficient as you move away from that point.
Limitations may be due to blower motors stalling, or not producing enough pressure to actually operate the diffusers, or there not being enough speed to cause surface or draught-tube aerators to actually entrain oxygen.
But aeration is important! What’s having too much of it do?
Over-aeration typically causes floc-shearing; the aeration energy applied causes the bacterial floc to be ripped apart, making smaller, far less settleable flocs. This, in turn, impacts setteability. Excess mixing energy will do the same thing. This is compounded by the flocs typically being weaker at low loading.
If you don’t get floc shearing, over-aeration can significantly reduce anoxic capacity as well. You’ll see either the DO in your aeration tank shoot way up above 2 mg/L, or the DO in your anoxic zone (or phase) increase above about 0.5 mg/L. As there’s less BOD around to actually remove this oxygen you end up losing de-nitrification without extra feeding. This can have an effect on both nitrate concentrations and pH.
So how do we fix that?
We’ll get into that in the next section, don’t worry.
Man. Ok, what else is going to go bad?
I mentioned that floc shearing can cause settlement issues, so you’ll also run into issues with your clarifier.
But it’s not just floc-shear induced.
Most gravity sedimentation clarifiers are designed to operate at a solids loading rate that results in hindered settlement. Hindered settlement is where there are sufficient solids present to cause particles to settle together. The discrete settlement is where particles just settle under their own mass. Hindered settlement is what forms a noticeable sludge ‘blanket’ in the base of the clarifier.
Hindered settlement is far more efficient than discrete settlement and can achieve far lower TSS concentrations in final effluent. Most facilities are designed based upon MLSS concentrations of 2,000 and 5,000mg/L (often around 2,500 – 3,500mg/L, although some can be higher or lower); if your MLSS drops too far below the lower design figure you probably start to have turbidity issues.
Filters can compensate for this, but you’ll find yourself backwashing more, and you may use more chlorine too to compensate for the higher solids loading.
You can look at lengthening out sludge age to maintain around your design MLSS to keep hindered settlement going, but you do need to be careful of filamentous bacteria growth, which will also impact settlement.
Ugh. Sounds like that might be hard to balance
Yeah, keeping your clarifier going at low flows is typically a balancing act between maintaining hindered settling and keeping a short enough sludge age to maintain good sludge health, and finding the sweet spot can take some trial-and-error.
What if I have an MBR? I won’t have to worry about settlement then.
That’s partially true.
Membrane Bioreactors can be a little more resilient to low loading as they are not reliant on your sludge settling well. However, they do tend to be more sensitive to the actual physical properties of the sludge.
Low loading rates can cause the bacteria to emit sticky substances (“Extracellular Polymeric Substances or EPS”), natural polymers that help rope them together. They can be useful, but too much of them in an MBR tends to plug the membranes up rapidly.
Filamentous bacteria can also cause problems with membrane plugging, causing rapid transmembrane pressure loss when in production.
So you do need to balance your sludge age carefully with an MBR too, otherwise, you’ll find yourself having to clean excessively.
How about SBR’s?
SBR’s can actually handle low flow reasonably well so long as they’re designed to do so. Quite often plants in areas that actually do encounter significant seasonal variations are SBR’s with reactors that can be shut-down and started up to meet seasonal loads.
Note that reactor shut-down and start-up tends to be a multi-month process; you can’t just turn it off for the week and expect it to come back strong; you essentially need to recommission the offline reactor once you want it back online.
So if you’re only going to be at low flow for a month or two, it might be better to adjust aeration rates and feed the process, which we’ll go over in the next sections; shut-down and start-up of a reactor is likely to take far too long. If it’s over the course of 4 or more months, then shutting down a reactor becomes far more beneficial, as you reduce the cost of process feeding and aeration.
Turning parts of the system off works best for systems with multiple reactors, preferably multiple sequential reactors (i.e. systems with 2 or more decanter tanks). Systems that have multiple reactors with a single decanter (typically constant-feed, intermittent decant systems) will work less well, but you may still be able to de-commission the first tank in the train (assuming your process configuration lets you do so).
For sequential reactor systems, if you’re lucky they’re designed to be turned do, so it might be as simple as hitting a button and maybe changing a few valves allowing one of the reactors to be shut-down. If it’s not designed to do so, so long as you have sufficient process control, you should still be able to turn some of the reactors off.
If it’s a two-reactor system you’ll then be running the process as a constant inlet, intermittent decant system. This can cause problems with effluent short-circuiting, unless you have appropriate baffling to prevent this.
If you have more than two reactors, and you’re dropping a number of reactors off of the end; make sure your process can be re-timed to ensure it works properly.
Managing a Plant at low flow
Okay, so, we’ve gone over the problems, now let’s work out how to make your plant work.
How, do we manage it?
Firstly, you may need to make a strategic decision and consider whether you truly need to run your wastewater plant.
If your flows are low enough, it may be a simpler option to truck your wastewater away, allowing you to either shut down your plant, or to only provide partial treatment before removal.
This, naturally, is only viable if you have a wastewater treatment plant nearby that can accept the waste, and if trucking costs are reasonable.
This option will cause issues when you go to re-start your plant; if you’re keeping it ticking over, you can re-start it by ramping up food dosing appropriately (more on that at (link)), if you shut it down completely you’ll need to re-seed the process from scratch, which can be harder.
So if tinkering is too expensive, what can we do if we still don’t think running the plant is worth the effort?
If you can’t tanker your waste away, you need to check if you have sufficient storage capacity.
If you have significantly sized storages, say for wet weather events, you might be able to store effluent for a period of time, and re-treat it later when your load increases and you can run your plant stably.
This, however, can be dangerous when you don’t know how long you’ll be offline for, as you may end up filling your storage and have no options about where you put your effluent. Storage of raw sewage can also create odour issues as the effluent turns anaerobic. At this point, it may be better to partially treat your effluent through your process and store the off-specification effluent for later treatment.
Okay, so if we can’t truck it away, and we can’t store it, we need to treat it, right?
Okay, we need to slow this thing down, what’s the deal?
To do this, you need both turn your plant equipment down as far as it can while still operating.
It’s likely that this won’t be enough to allow it to function, so you’ll also need to feed your process using an external food source. This allows you to get it at least mostly stable, so you won’t get knocked on the head by a flow or loading surge.
Okay, food and equipment turn down. Where do we start?
Let’s start with turning your gear down.
This involves adjusting aeration parameters, recycle rates, chemical dosing rates and wasting rates; these are the main parameters you’ll have control over as a process operator or owner.
Let’s start with Aeration.
Plants with cyclic aeration, such as Sequencing Batch Reactors (SBR) or intermittent aeration tank (IAT) systems can be turned down by simply reducing the aeration period, however sufficient aeration needs to be provided to prevent the plant from going anaerobic, which can happen if your Anoxic period is too long.
How do I work that out?
The simplest way is, if your sludge starts to smell or goes black, you’re going anaerobic and you need more air.
You may also see your Ammonia or BOD concentrations increase in your effluent; that’s also a sign you’ve gone too low in your aeration.
For cyclic systems, it’s preferable to stay above about 15 minutes of aeration per hour. You can go less than this if you’re confident you’re not going anaerobic. Not going anaerobic is important as anaerobic pockets in your plant can cause issues with bacterial growth, foaming and other biology issues.
If you’re operating an SBR having an aeration phase prior to settlement can help ensure you don’t go anaerobic by the end of your decant phase.
What if we don’t have cyclic aeration?
Many treatment systems won’t have cyclic aeration. This includes processes like the Modified Ludzak Ettinger (MLE), Modified University of Capetown (MUCT) or some forms of carousel oxidation ditch (among others).
Ideally, these systems should be designed with variable speed blowers allowing for aeration input to be increased or decreased. These variable speed systems can have lower limits for turn-down however, as there can be issues with motors stalling or maintaining cooling.
The exact method and how well the process can be turned down will vary based upon how the aeration system is designed.
If the blowers aerate directly, with speed controlled either by operator input or a DO sensor, you’ll be limited by the motor turn-down. This will typically result in your blowers running at their lowest speed, and possibly your DO concentration sitting higher than set-point. The high concentration itself is not an immediate cause for concern, but does indicate you may have some other issues (e.g. reduced anoxic capacity or floc shear).
If you are having problems with floc-shear or anoxic capacity it may be necessary to look at whether a temporary timer can be installed to further turn-down aeration capacity.
If your blowers are controlled based on manifold pressure, with aeration rates controlled via valves, you may have an easier time.
The system will naturally shut down the blowers and diffusers to try to maintain the target DO set-point, and for this, you need to ensure that you’re actually maintaining mixing within the mixed liquor; if not it may settle and go anaerobic at the base of the tank (since the DO sensor is at the top it’s possible to have an aerobic surface and an anaerobic base without adequate mixing).
In this situation, if you have the capacity to do so, turning off some of the diffuser rails should help you reduce overall aeration, just make sure you maintain adequate mixing. Also make sure you’re not pushing too much air through the remaining diffusers, as this can damage them, which will cause problems down the track.
Finally, implementing a timer may also help reduce aeration capacity.
Ok, aeration is sorted. What’s next?
Next, you want to look at turning down your recycle rates.
First to look at is the Return Activated Sludge (RAS) recycle, which is common to all Activated Sludge (and quite a few fixed film) processes.
You typically want to pace your RAS rate at somewhere between 0.5 and 1.5 times the raw inflow, and generally, you’ll have trouble keeping your RAS rate down at low flows.
What’s pushing too much RAS do?
Over-RAS’ing tends to drain the sludge from the clarifier, which interferes with the hindered settling process I mentioned earlier; essentially it drains the blanket out of the clarifier, reducing clarification, which can impact final effluent TSS concentrations and turbidity.
Try to get the RAS as close to the design ratio as possible, and if you can, monitor your sludge blanket height (or depth if you will), as this gives you an idea as to how appropriate your RAS settings are.
Check your clarifier performance; the combination of lower flows and lower sludge loading may result in you getting reasonable performance without hindered settling… you’ll still probably have increased effluent TSS, but it may not push you out of spec.
What about the other recycles?
If you have other process recycles (such as the anoxic recycle in the MLE) they should also be set as close to their design ratios as possible; rates too high tend to cause issues in the process; e.g. too high an A-recycle rate can recycle excess oxygen to the anoxic zone, which reduces the anoxic capacity of the plant as the biomass will utilise oxygen before nitrate. Anoxic zones can be even more sensitive to oxygen inhibition at low flows as the amount of BOD available is reduced.
Worse comes to worst, intermittent recycling can help reduce flows; once again implementing a timer can come in handy.
Okay, we’ve sorted out our recycles, what’s next?
If you have process mixers you should also check them too. As we discussed, too much mixing energy can cause floc shear, so if they can be turned down via VSD then turn them down.
What about our chemical dosing?
Yeah, chemical dosing also needs to be adjusted too.
Chlorine dosing rates should drop due to the lower flow, as should dose rates for phosphorus coagulant chemicals such as ACH or Alum and any pH or alkalinity control you might utilise (i.e. Lime, Caustic or Soda Ash dosing).
Our dosing is flow paced, will that be an issue?
So long as you have the turn-down rate available you should be ok. Depending upon your pump branding dosing pumps typically have really good turn-down rates.
Often getting to the low flows with your dosing gear isn’t the issue, it’s keeping the pumps going at low flow rates.
Some chemicals (I’m looking at you, sodium hypochlorite) can tend to gas-out, which can gas-lock diaphragm dosing pumps. This normally happens when the pump sits idle (which happens more at low flows) but can happen at very low flows.
You can also run into issues with flowmeters and flow switches if you’re running well below what their alarm points have been set.
If your pumps dose without any control input, then you’ll need to work out new dose rates based upon the lower flows and adjust them accordingly.
I assume too much chemical dosing is bad then?
Like everything else we’ve discussed in this, over-dosing chemicals can cause problems with your process, with different chemicals causing different problems.
Over-dosing ACH or Alum can cause problems with nutrient availability (the chemical locks up all the phosphorus needed for bacterial growth) sludge settlement (some ACH or Alum will help sludge settlement, too much will make it float) and pH (especially with Alum, which has significant alkalinity demand).
Over-dosing chlorine can lead to non-compliance with your license limits, as most licenses have a limit on chlorine. This isn’t limited to sodium hypochlorite systems; low flows through older tablet dosing systems can cause problems with chlorine dosing if the tablet chlorinator is not configured correctly, as you can get tablets dissolving into little to no wastewater flow causing a slug of high chlorine to pass through the system when flow resumes.
You can also have issues with significant retention times in your dosing race, which can cause problems with disinfection due to chlorine dissipation and stagnation. Over-dosing pH chemicals usually lead to the pH increasing beyond the license limit, which primarily impacts disinfection performance.
Okay, cool, what’s next?
Finally, you need to look at your WAS rates. Getting your WAS right will help a lot when feeding your treatment plant.
You can either increase your WAS to maintain your treatment plant’s design F:M ratio, or reduce your WAS to lengthen your sludge age help ensure there is sufficient MLSS to maintain good clarification (although this can be risky as you can end up with filamentous growth and sludge bulking due to very low F:M at longer sludge age).
The actual answer is probably somewhere in between these two, and I’ll get into the reasons as we discuss process feeding.
Okay, so now that you’ve adjusted your aeration, your recycles, your WAS rate, any mixers, and your chemical dosing systems; you need to start working out how to feed your treatment plant. We discussed how to work out how much food to add previously, now let’s go over how and where you should be adding it.
So, remind me, why do we need to feed the STP?
You’re likely to be limited in how far you can actually turn your plant down. A STP typically has a minimum loading, below which it won’t be stable; if your raw sewage loading isn’t at or above the minimum load it’s likely you’ll need to feed the treatment plant with an artificial food source to maintain stability. Process instability can result in intermittent (or constant) failure to comply with effluent specifications, poor response to shock or toxic loads, etc.
Ok, so we need to feed the plant to keep it operating properly.
Let’s go through the chemicals you’ll probably need to look at dosing so we have an idea of what we’re getting into. We ran through these in a flyer we released (link), so I won’t go into depth here.
There are 3 major nutrients your treatment plant needs, that’s carbon (carbohydrates, BOD) Nitrogen (Total Nitrogen, ammonia) and Phosphorus (Total Phosphorus, orthophosphate). There is also a range of micro-nutrients, including iron, calcium and magnesium that are also important, but are typically at far lower doses than CNP, and you might just get away with what’s in your raw sewage (or not, as the case may be).
Carbohydrates are generally fed as a simple carbon source. Sugar, sucrose solutions and molasses are common, but you can also dose sodium acetate or methanol, which are less common; sodium acetate is expensive and methanol requires specialised storage as is flammable.
Granular sugar can be pretty useful since it’s simple to make up and readily available. Sucrose solutions are also pretty common, both as liquid sugar for food manufacture and sucrose solutions for de-nitrification support. Both Sucrose (at a concentration of 60%) and Molasses are thick, viscous material; they can be dosed directly with the right equipment, but may require some dilution if you don’t have an appropriately designed dosing system. In this case, making up a granular sugar solution may be more viable.
Nitrogen can be from Urea, ammonium solutions, or nitric acid, with Urea generally being the easiest to handle. Solid urea can be dissolved in a bucket and dosed, and you can also get liquid solutions. Ammonia solutions can be applied but can be hazardous to handle, so you need to take appropriate care. Liquid solutions should be able to be dosed simply with dosing pumps.
Phosphorus can be from a phosphate based fertilizer, like Mono-ammonium phosphate or di-ammonium phosphate, or Phosphoric acid. The fertilisers are probably easier to handle, but do add in additional nitrogen, which needs to be taken into account when considering how much urea or ammonium you’re adding. Phosphoric acid should be able to be dosed simply with a dosing pump. Again, take due care with safety and set-up.
Finally, it’s a good idea to add something like blood-and-bone and kelp concentrate, which helps with trace elements like calcium and iron. Be careful when adding this though, as it can add in fats and oils, which can, in turn, cause filamentous bacteria growth, which causes bad foaming and settlement issues.
To work out how much you need to dose, refer to the flyer I mentioned previously (link), it has good info on how to estimate the volume you’ll need, and how much your chemicals might end up costing.
Ok, I’ve got what I need. How do I go about feeding it?
How you feed your process will vary based upon the form of the material you have, whether you have access to dosing equipment, and how often your operators are on site.
If you’re using solid solutions, it’s generally better to make them into liquid solutions and dose them with a dosing pump. It’s best to keep the solutions separate since you can get biological reactions within your storages that tend to make them stink.
You can dose dry chemicals directly into the plant if they are readily dissolvable; if they’re not, it’s best not to, as the chemicals will tend to just drop to the base of the tank and slowly diffuse, which is not good for mixing and tends to just gunk up your tank over time.
Also, be careful of dust generation.
If you’re using liquid chemicals, I’m almost always going to recommend using a dosing pump. You can pour them into the tank too, but having a pump gives you some other options regarding scheduling and pacing.
Okay, when do I dose?
Ideally, it best to just dose the plant slowly over the course of the day. That helps even out loading and stops you from shock loading your plant, which can cause poor performance.
This is why using dosing pumps can be really beneficial in this process since you can set a dosing rate and walk away.
You can hand dose, but if you do so try to spread the doses out over the course of the day. Hand dosing once or twice a day tends to cause loading surges, which can result in poor treatment performance.
Right, is there anywhere specific I should be looking at dosing, or can I put everything in the pumping station that feeds the plant?
Generally speaking, dosing directly into the bioreactors tends to be the simplest method. This avoids any loss of material in up-stream processes.
If you have a primary sedimentation, make sure you dose the feed after the sedimentation chamber, otherwise, you run the risk of settling out or digesting some of the chemicals you’ve just added, which is a waste.
If you don’t have primary sedimentation, you can dose into the headworks. This can help if you have multiple bioreactors (like an SBR) since you can use the splitter system for the inlet works to distribute the feed. You do need to make sure you have flow to push the material into the bioreactors however. If you don’t have relatively consistent flow you may need to look at a method to create it (i.e. recycling effluent through the headworks, or just using a hose). Without consistent flow coming in you run the risk of surging your tanks in a similar manner to just hand dosing since the feed will only come in when raw sewage does, and since you’re at low flow, that’s probably going to be infrequent.
A flow buffer tank can help with metering feed into the process, but again flow needs to be taken into account. If the balance tank accepts a recycle from elsewhere in the process, like the RAS, you should be fine. Lots of small treatment plants have activated balance tanks (i.e. the RAS is returned to the tank), which should work quite well, and will help significantly with evening out doses (if you’re dosing by hand).
You can also look at dosing to target specific problems, such as dosing extra sugar during the anoxic phase or zone to assist with de-nitrification. This does, however, need some fairly specific process consideration and should probably be done in consultation with an expert or experienced operator.
Any other pointers?
Wherever you end up dosing, try to make sure it’s in a mixed area; the aeration tank near the plant inlet, or into a mixed anoxic zone are both good choices.
If you have cyclic aeration, dose during aeration to ensure good mixing occurs (unless you have de-nitrification issues, then dose some of the sugar during the anoxic phase) don’t dose in a settlement phase since it won’t mix properly. If you have a mixed anoxic phase then you can dose whenever.
As I mentioned, the ideal method is to pump the feed solutions into your process slowly, over the course of a day. This can be via a spare dosing pump that’s temporarily set up to dose into a mixing zone, or even a cheap submersible pump dropped into your mixing tank (which might work better if you’re dosing into an inlet works, since you can bulk the feed solution out with a lot of water to compensate for flow); just make sure the solution volume and pump are sized appropriately.
If you can’t meter the solutions into your process slowly, try to add chemicals in multiple doses over the course of the day; add some in the morning, some around lunch, and then some before you go home. It’s not as good, as a consistent dose, but it’s better than just dumping chemicals in once a day.
Okay, so, we’re sorted then?
Yes, but also no.
Remember we discussed the WAS rates before, and I mentioned they’re probably going to be a bit of a constant battle? Well, the main issue you’ll typically run into when trying to keep a plant alive via external feeding is sludge bulking and poor settlement.
This is driven mainly by the fact that filamentous bacteria thrive in a very low F:M environment, and simple sugars also tend to favour filamentous growth, as they can cover a far greater area than floc formers, helping them grab food more effectively.
Sounds bad, how do we fix that?
I think I mentioned it before, but filamentous bacteria growth is typically managed by adjusting your sludge age and F:M ratio, your WAS rates, and how much sugar you’re dosing.
If you can, it’s beneficial to get a microscopic inspection of the activated sludge conducted; good inspection can tell you what bacteria is present and what factors favour its growth (e.g. high sugar loads, low F:M, anaerobic conditions, etc.) which can be of significant benefit in determining troubleshooting steps.
If you run into settlement issues make sure you get a MLSS (mixed liquor suspended solids) test done; don’t just rely on SSV (settled sludge volume). Knowing your MLSS and your SSV allows you to calculate the SVI (sludge volume index), and gives you a good idea of how much sludge you have if you do need to start wasting heavily (waste too much and you’ll run into issues like loss of nitrification).
Increasing WAS lowers the sludge age which both helps to select against filamentous bacteria (basically makes them less likely to grow) and helps to clear the bad biomass out of the system.
Lowering the sugar dosing also helps to reduce the competitive advantage the filamentous bacteria have, but you have to be careful you don’t lower the F:M too far.
Usually you’ll do one of these at a time, and whichever you chose is often determined by other factors, such as chemical costs or sludge storage availability.
Depending upon how your process is going you will sometimes need to waste aggressively to get bulking under control. This can cause problems though, since you might be running at very low biomass concentration already, and may not be able to waste much more. You may also have storage limitations that prevent you from wasting as aggressively as you need to.
Yeah, I can’t waste any more. What now?
There are other methods you can use to boost settlement, but these tend to be a bit more involved than altering F:M and the sludge age and can have major downsides if unsuccessful.
Ok, hit me.
First up, you look at using a coagulant to help improve sludge settleability. This can be relatively simple to implement if you’re already dosing for phosphorus removal, but you need to be careful with how much you dose; if you over-dose there’s a good chance it may cause the sludge to float instead of improving settlement.
Also, if you have low flows and not much phosphorus around, dosing extra coagulant can further damage your biomass by locking up all the available phosphorus, which it needs to actually live. This causes poor treatment and can lead to more sludge bulking.
If you want to try dosing a coagulant to improve the settlement, make sure you do bucket tests first.
It’s just a simple test to see how your sludge will react to the dose.
Take some sludge and mix in various amounts of coagulant. Check the results using an SSV test and visual observation of the sludge. This helps you see if you’re going to cause settlement issues at the chosen dosing rate. Failure will normally be pretty obvious; it’ll do nothing or it’ll float. If you don’t get a good dose rate, try again.
The dose rate for ACH is far lower than Alum. Alum tends to create more sludge, so it’s actually better for this application. You could also use Ferric Chloride.
The equipment itself doesn’t have to be complicated; you can even do it in an SSV cylinder, which actually helps since you need to do an SSV anyway. Just mix the coagulant in by pouring it between two cylinders and leave it to sit. Don’t stir it in a jar test rig (if you’re lucky enough to have one) as the slow stirring that’s typically used for coagulation testing can significantly improve the settlement since it allows water to be released from the sludge.
If you are using Alum or Ferric you do need to be careful with your pH, as both of these chemicals have high alkalinity demand; check the pH of your bucket test since you may need to add something to increase it (e.g. Lime, Soda Ash, Sodium hydroxide, etc.). ACH tends to have a lower pH impact so you might get away without alkalinity dosing, but you should still check the bucket test, just to be sure.
I’m not sure coagulation is the right fit for our plant, any other options?
The next option is to try to make your sludge heavier so it settles faster. We usually do this by dosing powdered clay material, typically powdered Zeolite or Bentonite.
If dosed properly the particles are absorbed by the sludge floc, which increases their weight, improving settleablity.
Sounds good, why didn’t we start with that?
Two reasons; the actual mass you need can be confronting (typically 1/3rd of your activated sludge by weight, which for large reactors can be a significant amount), and adding it to your tank can be difficult to get right. If you under-dose or don’t add it properly there’s a good chance it’ll just do nothing.
The method used to dose needs to ensure that the clay is evenly mixed into the activated sludge, so it gets included into the floc; if you dump a bag or two in the same spot off of one of your access walkways the material will just sink to the bottom of the tank sitting there.
The best methods are to broadcast the material over the surface of the bioreactor, or to dose it directly into the RAS line at a constant rate.
Tell me how that works.
To broadcast the material it’s best to use a wet solution or slurry, as this helps with both dust control and application. The slurry can then be sprayed in with a hose, though you’ll also need a pump that can handle a slurry solution and a mixer to keep the material in solution before its pumped. Spray evenly as you can over the surface of the tank and let it mix into the sludge.
Broadcasting gives good mixing with the sludge, as the material sinks into the sludge mass allowing for relatively even absorption. You need to dose while mixing, so into an aerated or mixed phase if you use cyclic aeration, or an aerated or mixed area if you have constant aeration/mixing. Aerated phases or zones are generally better than mixed, as the mixing caused by the aeration system tends to be more intense than that caused by a mixer.
Alternatively, you can dose into the RAS line, although this will need equipment and a dosing point. Dosing into the RAS line also ensures good mixing (assuming flow is turbulent; having a bunch of fittings after the dosing point also helps get good mixing), and if you have the equipment to do so it can be easier than broadcasting.
If you are dosing into the RAS it needs to be over the course of an entire day as this will ensure good contact between a large volume of and the material. If you throw it in in an hour only a portion of your sludge will actually be dosed, so you’ll have a bit that settles really well, but most of it won’t.
You also need to maintain a ‘maintenance dose’, to offset any material being taken out by sludge wasting. For this you dose about 1/3rd of the mass of sludge wasted is a good number, and this can be done on a weekly (or so) basis.
Any more we can look at?
Finally, there’s what’s probably the most risky option; chlorination of the activated sludge.
This works by killing any bacteria that isn’t incorporated into a sufficiently large floc structure, and can be effective on long, stringy filamentous bacteria, due to their high surface area (which is what helps them out-compete floc formers for food at low loads).
Chlorine would typically be applied to the RAS line and can work but it is very risky and would usually only be used as a last resort.
If you mess up the dose rate, at best it’ll have no effect and at worst, you’ll kill your sludge and need to re-seed the process. Because of the risk, it should really only be considered a last-resort, and if you are considering it, get specialist advice.
Should you need further information, please contact us
Email: firstname.lastname@example.org Or Call 1800 620 690
Click to download the whole document: Talks with Terry – Running a Process Slowly