Washable Filters to Reduce Total Dissolved Solids (TSS) in Produced Water

Total Dissolved Solids cannot be removed by filtration as the particle size is ionic.

Total Settlable Solids [TSS] can be removed by a variety of means of filtration, however the units of measurement are not correct for TSS. Further, there is no flowrate to look at; is it 1 l/day or 300m3/h, for example?

Please clarify.

Sorry,

Total Suspended Solids (not dissolved). and a flow rate is 1000 m3/hr.

Thanks for correction.

That is a very different game from TDS - suspended solids can be settled out and in the end filtered - one of the guys will know this procedure/equipment well.

Look for cartridge filters or depth filters. They go down to a micron, or less. Don't know about washable. I don't have time to do the work for you.

What's the largest allowable particle size? If it's 5μm a cartridge filter of that size will do the job. Select the filter based upon maximum allowable particle size.

1000 m3/hr and 140 mg/liter is a good lot to filter.

Dead-end filters, like cartridges, for a 1000 m3/hr flow of oily produced water sounds like a fool's errand. You need to look for a crossflow filter.

You need a bank of these succers

These are filter packs with backwash capability. This bank is about 1/5th of the size you need based on your flow estimate.

Each unit can be backwashed while the others are forward flow. Provides continuous throughput.

Not cheap!!

PWSlack suggests your units are wrong and I concur. Are you referring to nephelometric turbidity unts (NTU) as in water turbidity or mg/L as in TDS. If 160 mg/L is really NTU your water will look like mud. You can get coarse filters such as JUDO stainless (it is on google) that require very little backwash water and can even be automated. To get to 20 micron will require pleated washable filters. You could go from 50 to 20 to 5 micron pleated washable filters. Anything 1 micron or lower cannot be cleaned. It may be prudent to recheck the quality and units to confirm what you are dealing.

By the way you can clean the pleated filters with a 3% solution of peroxide soaking for a few hours and then rinsing. You will only be able to do this for a limited amount of washing before you must replace cartridge with new. It will still be a substantial footprint for the flow rate.

Nickel 20 micron microsieves are available from Veconic. With a 1 m/s crossflow, 4 micron particles (yeast) were retained using such a microsieve in a small experiment by van Dinther et al. published earlier this year. They demonstrated high flux without clogging. The crossflow velocity, and therefore the retention and effective pore size, can be much greater than that for your application. Here is a possible configuration for handling the huge oily gritty flow you are dealing with: http://www.freepatentsonline.com/7757866.pdf.

OP has yet to correct units and/or quantifying quality. If the water is mud, he ought to be looking for stilling ponds or settling clarifiers and coagulant agents. I was uncertain if he/she is dealing with waste water and what is the target discharge. Is oil present? Is this supposed to be water treatment for human consumption? If so, OP will need a better source. A turbidity of 160 NTU is very high ( I assume OP made a mistake and reported mg/L). If it is to be used for water consumption, OP needs to look for better source protected areas. I suspect no filter will work if the quality is so poor. Your links were poor in that they only provided an abstract with a fee to get the meat. The type of filter will depend a lot on OP clarifying the units and confirming quality.

Look at using a bank of media filters in pressure vessels, a.k.a. "sand filters". They will need to be engineered, as an off-the-peg solution is unlikely to fit.

To add to others' comments, this is a much bigger job than you seem to think. If you remove 150mg/l, that's 150kg/h solids, and the filtered sludge might be at only ~ 10%, so that's 1500kg/h sludge. A replaceable-element cartridge type is a non-starter.

As it's produced water some of the solids likely to be oil, so you might be able to use a coalescer of some sort (eg tilted plate separator), which takes it out as liquid oil rather than in a filter medium. I doubt if you'll find a roughing filter (eg automatic backwashing type) to take out stuff mostly < 20mμ.

What temperature is it? Last time I dealt with produced water (North Sea) it was 107°C.

As PW Slack says, it looks like a deep-bed filter job (say 1m deep sand, 1mm sand size). Depending on any pretreatment, filtration velocity in region of 5 - 10m/h, so it's a lot of filtration area. Likely to need chemical dosing - ferric chloride and maybe polyelectrolyte.

Codey