5 replies to this topic

[matt719]

Hello everyone,

I should probably preface this by acknowledging that this not a large scale application, but I would appreciate the insight of an engineer. I recently graduated from a large university with a degree in chemistry. I'm working for the university's EHS department this summer doing a project on hazardous waste management. The university gets large amounts of dilute aqueous ethidium bromide waste from research labs. For those unfamiliar, ethidium bromide is used as a stain in gel electrophoresis and, as it complexes with DNA, is a potent mutagen. We want to remove this trace ethidium bromide from many 20L waste containers and safely dispose of the water/buffer down the drain in order to minimize waste disposal costs. This is traditionally done using activated carbon, where the requisite amount is added to the waste, stirred for ~24 hours to ensure complete adsorption, and filtered.

I want to employ activated carbon to capture ethidium bromide, but I would prefer a continuous process as opposed to the batch process that has been heavily documented/marketed. I'm trying to determine if I can make some kind of packed-column activated carbon filter to remove ethidium bromide from waste water. The concept is to be very low cost. My idea is to have two ~25L vessels (plastic carboys, say), one on top of the other, with a filter in between. The waste water will feed through the filter by gravity and the clean water will come out into the lower vessel, where it will be checked by UV light for remaining ethidium bromide before it is disposed of. I imagine if my filter is large enough and packed tight enough, this would work.

Does anyone have any experience filtering ethidium bromide, or have any thoughts? I imagine it would be a similar process for any trace organic chemical.

Thank you!
Matt

[Erwin APRIANDI]

Hi matt,

Are you sure ethidium bromide can be adsorp by activated carbon, and how effective it is.
I have made a small waste water treatment in semi continuous process for medical waste.

The problem in employing this kind of operation (semi continuous) for waste water operation is that
you should have a preliminary treatment in order to remove particulate from the waste, this ussually can be done by filtration using sand media filter.

and the semi continuous operation can be done by minimum having 2 set of filtration in series, so when one set of filtration is being used, the other one will undergoes back wash and regeneration.

I could help you in finding the right design, but I have to know the exact composition of waste that you have and what will be the target of effluent after the treatment

[kkala]

It seems that ethidium bromide absorption by activated carbon is rather slow, seeing that its solution requires about 24 hr residence time in stirred tanks. Assuming same residence time in the carbon filter, its volume (not including "solid" volume of activated carbon, expected to be a fraction of the total) would be for treating 10 bottles per 24 hours (flowrate 10*20/24=8.3 l/h, say 10 l/h to compensate downtime):
10 l/h*24 h=240 l.
Nevertheless needed residence time may be different to the one indicated by agitated vessels, but only experimentation can show it, in a smaller scale. So better to seek first results by doing measurements in the lab, using same configuration in a small scale (glass bottle, carbon filter, glass bottle). Flow would be quite small, manually controlled by some kind of lab "valve".
I wonder whether even flow distribution through the activated carbon bed is significant in this case, since flow channeling cannot be predicted by above lab tests. If it is significant, carbon adsorption column may require optimum length to diameter ratio to minimize channeling. But liquid velocities will be so low that channeling may not play a role (adsorption to be controlled by natural diffusion from liquid to carbon particles in the column). It is assumed so herebelow; however advice would be welcomed, having no experience on this matter.
Under this assumption, agitation would promote "transfer" of ethidium bromide to the carbon particles, which is realized in the traditional batch method of stirred tanks.
So the traditional method of ethidium bromide removal may have reason to be so. At any case a lab test (as mentioned above) is judged worth while.

Edited by kkala, 23 July 2011 - 12:42 AM.

[matt719]

Thank you both for your thoughtful responses.

Erwin,

Thanks for the tips about the preliminary treatment and backwash and regeneration. Unfortunately, it's very hard to know the "exact" composition of the waste. In truth, it will vary substantially from batch to batch, and it's impossible to know exactly what was put in the waste container. But, as the processes often involve buffers, a fair amount of salt should be expected and some type of preliminary treatment would most certainly increase the longevity of the carbon. In regards to regeneration, this is something we won't be doing. The facility is not technically allowed to treat waste, so we have to be careful about what we're doing. The idea is to capture the ethidium bromide using carbon and then dispose of the contaminated carbon as solid hazardous waste. This is more cost-effective, and arguably creates a smaller environmental footprint.

Kkala,

Let me see if I followed you. To have the same residence time in the semi-continuous process as the batch process, a flow rate of 10 liters/hour would be necessary. However, it is impossible to say whether it is simply a matter of residence time of the liquid with respect to the carbon; it's possible that there is something about the agitation that promotes adsorption that wouldn't be mimicked by the semi-continuous process and only experimentation will tell. In addition, it'd be smart to use a column that is long with respect to its width to minimize potential channeling. Is that right?


Matt

[kkala]

Following tries to clarify post of Jul 23rd by kkala.
To have the same residence time in the semi-continuous process as the batch process, a flow rate of 10 liters/hour would be necessary.
For 10 l/h of EBs (ethidium bromide solution) treated, 240 l of activated carbon is necessary (plus volume of "solid" carbon), on the condition that residence time of 24 h is required (of course according to stated assumptions).
Capacity of 10 l/h is arbitrary chosen, treatment of 5 l/h of EDs would require 120 l of activated carbon. It depends on how many EDs bottles should be treated per day (24 h), which you have to specify.
However, it is impossible to say whether it is simply a matter of residence time of the liquid with respect to the carbon;
Residence time of liquid = 24 h is a first assumption, for a feeling of the required quantity of activated carbon (indicative conclusion: this quantity may be rather big, so smaller scale lab tests are indicated in advance).
it's possible that there is something about the agitation that promotes adsorption that wouldn't be mimicked by the semi-continuous process and only experimentation will tell.
Yes. Due to very low velocities of EDs in the activated carbon bed of the "new" process, transfer of ED to carbon particle sphere may be by natural diffusion, which would be faster by agitation (this does not concern transfer of ED inside the pores of the carbon particle).
In addition, it'd be smart to use a column that is long with respect to its width to minimize potential channeling. Is that right?
If channeling plays a role, which may not be the case. In such small liquid velocities, activated carbon particles are assumed totally wetted. Local velocity variations (if any) may not play a role. We do not have gas counter flow here, so channeling is not the usual one.
Advice would be welcomed on the last two matters (agitation and channeling).
Consequently the traditional method may have reason to be so, although a lab test of the method under consideration may be worth while.

[Erwin APRIANDI]

For uncertain waste composition and flowrate also due to low, adsorption time

I suggest you keep the idea of having batch operation, for continuous operation
the main problem will be in case of your effluent to be discharge is out of spec
which is it most probably be happen in your case.

The uncertainty margin in waste water treatment is quite huge and even after doing 6
waste water treatment system design and construction, I even never use 10% as my OD margin since I have to made a process guarantee. My minimum margin for any design of waste water treatment is at 20% OD margin