Extreme Flow Variations & Process

[sjohns4]

I'm preparing a PER for a training facility that sees incredibly variable flows. The variation will happen weekly. There can be intermittent periods of extremely high flows (~140,000 GPD) flowed by periods of extremely low flows (~8,000 GPD), and various levels between. I'm using 350 mg/l BOD & TSS at the moment as the collection system is minimal and will be brand new. Actual water use and raw flow testing from a similar facility are in the process - regardless of actual flow and loadings, the same variability and flow swings will be there so I think the challenge will be similar.

We only need to meet a 30/30 effluent, no N or P limits, but we do have a turbidity limit of <2 NTU when using conventional treatment and filtration and a turbidity limit of <0.2 NTU when using an MBR.

In my mind, the flow is the easy part. I think the obvious solution would be multiple process trains. I'm not real confident on the bugs, solids, and overall process efficiency.

My first round of calcs on paper show I can use 2 trains and maintain parameters between operating the secondary process as a complete mix activated sludge plant during peak flows and an extended aeration type plant during low flows.
Complete Mix / Peak Flows:
2 basins in operation
Hydraulic Detention Time: 5.5 Hrs
MLSS: 4000 mg/L (1,086 pounds in basins)
Basin Loading: 95.25 # BOD/d/1000 cf
F:M - 0.51
SRT: 5.85 Days

Extended Aeration / Low Flows:
1 basin in operation
Hydraulic Detention Time: 48.8 Hrs (Long for EA...is that really a problem if we keep the DO right?)
MLSS: 5000 mg/L
Basin Loading: 10.74 # BOD/d/1000 cf
F:M - 0.05
SRT: 22.62 Days

Ramping down sounds simple enough: transfer mixed liquor from one plant to the next and waste sludge.
When everyone shows back up I'm showing aeration basin solids should be re-established in a about a day with no sludge wasting.

Assuming all of the above are correct and doable, how will the process react to the changes?

Thanks for any wisdom that can be provided.

Mike

 

[ashtree]

This is the type of situation that suits an SBR.
If these swings happen weekly than you would not change from one basin to two basin operation in that time.
If the variations are regular, lets say the low flow is always the weekend you would design a control system to deal with that. On those days you might aerate t a much lower rate , maybe little or no wasting, longer idle periods. The bugs will tolerate this provided the aeration roughly matches the load. If there is no load and the bugs have been aerobic they will sit there for some time living endogenously without a problem. Once the load comes back you need to start aerating proportionally at about the same time, unless you intend cycling them through a anaerobic or anoxic stage.
If the variations are not regular you might needs to include some sort of feed forward control that responds to a flow input, and increases aeration according to the amount received in the previous 4 hours say. You might also design the SBR so that the cycle times vary based on load.
There are lots of options for this.
If that peak load is way above the other flows an equalization basin might be able to be used to buffer the flow at peak times and release it to the process overtime.
Maybe a lagoon type single basin SBR might be suitable if but can be built large enough to cater for the peak flows and then gradually release the effluent over a longer period.
This approach may allow you to run longer aeration cycles, with fewer settle/decant periods in the day, thereby able to use smaller capacity aerators.

Regards
Ashtree
"Any water can be made potable if you filter it through enough money"

 

[bimr]

You would be best served with a system that includes a high performance influent screen, a RBC system, settling basin, and effluent filtration. You can install multiple RBC units and take some RBC units online and offline as required. RBC systems are frequently specified for resorts because it is possible to take modular units offline and RBC systems are simple to operate. You also will not have a large volume of mixed liquor. Systems with large volumes of mixed liquor will take longer to stabilize as the process conditions change.

RBC Process

Complete mix or extended aeration systems are not suitable for widely varying flows. It will take weeks for these systems to stabilize when significant changes occur in the influent.

Membrane systems and SBR's are too complex and small systems will be difficult to operate for the limited staffing that is typically available.

 

[ashtree]

There is another issue here which requires some focus.

Although you do not have a N or P limit and you say that you only have to meet a 30/30 requirement the 2ntu maximum turbidity potentially puts a fairly tough limit on your suspended solids. I am guessing that's why you are thinking about filtration and or an MBR.
2ntu is probably no more than 4 to 5 mg/l suspended solids in reality and maybe even less. Media filtration might be able to do that for you but you will have to manage the process ahead of the filters very well to achieve reliability. The filters will have to be well matched to the load fluctuations, hydraulically and quality wise that come with wastewater treatment and you may need an alternative disposal or treatment option when the quality is too bad to be filtered successfully.

Regards
Ashtree
"Any water can be made potable if you filter it through enough money"

 

[sjohns4]

Yes, to about everything that was said.

My knee jerk was a re-circulating sand filter or Orenco type plant. I used to put those in state parks for this exact reason. I knew they worked, but couldn't tell you why they could accommodate such varying loadings. The downside with either of those is footprint; we're getting a bit large at 140k GPD. I also see SBR's used around here for small communities, but the flow variations they were selected to handle is I/I, so it's all hydraulic load rather than mass load.

I really believed my suspended sludge scenario presented would be problematic, but here again couldn't really say why.

Filtering or MBR is a regulatory requirement for reuse. I completely agree, by the time you get down to 2 NTU you're going to be single digit TSS. My speculation was the filters will be doing a lot of the work, possibly with a coagulant.

Since flows are varying basically by week (week-long training sessions), an EQ tank for buffering weeks' worth of flows get pretty large. Volume of tankage aside, I wasn't sure how long was practical to hold raw wastewater in an EQ tank, even if aerated?

Thanks,

Mike

 

[bimr]

I would seriously look into RBC's. The start-up time is 1-2 weeks.

Most people are using cloth disk filters on the effluent side.

Facilities with high I/I utilize what are called excess flow treatment systems. If the flow exceeds the capacity of the treatment plant, the flow is diverted to a holding tank. Once the flow decreases, the contents of the holding tank are pumped back to the treatment system.

 

[sjohns4]

I took a look into RBC's, they do seem attractive. A lot of attached growth in a small footprint.
Figured an Aqua-disk or similar on the effluent side.

One of the membrane suppliers claimed they could provide a system to meet our needs. A couple things didn't quite smell right to me, care to sanity check?

They're proposing an MLE process with upfront anoxic to help keep pH from taking a dive when alkalinity is consumed during low flows and extended detention times when nitrification is occurring. I get the alkalinity consumption and recovery, but if we're just trying to remove BOD does it matter if pH limits nitrification?

The other part I'm questioning is that we'd have extra membrane capacity to account for biological process upsets. Basically, when the process is upset, the membranes would be fouling much quicker, so we'd need to allow for additional cleanings. Expense and pain of additional cleanings aside, does this sound acceptable? It partially makes sense in that problems in the aeration basins become tangible when solids don't want to settle in the clarifiers and since we don't need to settle solids no problem, but it also seems with a UF membrane you could produce clear, low NTU water but still bust BOD due to dissolved organics?

I really don't like membranes for this application if we can avoid it, just due to O&M requirements, but for the sake of a complete evaluation...

Thanks,

Mike

 

[ashtree]

The pH change may be sufficient to impact on the other biological processes especially if the change is significant and rapid. Whether or not nitrification and subsequent denitrification is occurring will also impact on DO and aeration requirements.

Membranes may allow the production of clear low ntu high organic water but such scenarios will likely result in high levels of membrane fouling.

Regards
Ashtree
"Any water can be made potable if you filter it through enough money"