Freeze proofing an oil-water separator 1

[atrizzy]

What are the best ways to freeze proof an above-ground API style OWS?
Are agitators an option with the gravity separation mechanism also at play?
In place heaters? Should I insulate the sides of the thign and provide removable insulated lids?

Any advice is appreciated, and thank you in advance.

 

[CarlB]

What size is it, and what are the design winter temperatures?
I would think insulate at a minimum, including exposed piping.
Electric heat trace on piping. Tank itself could also be heated with heat trace placed around tank prior to spray foam insulating. I've seen that approach for sewage holding tanks in an arctic village.

 

[atrizzy]

Size is up in the air still, but probably somewhere in the range of 6' wide, 3-4' deep, 30' long.
Design witner temp is -40C. Piping approaching the OWS is all heat traced.

The tank may be cast-in-place concrete, so heat tracing the exterior might be inefficient.

Thanks for your input, we're really in the early phases here, trying to identify potential issues.

 

[dhengr]

Atrizzy:
I wouldn’t think that agitators and gravity separation mix (is that the right word?) very well in a oil-water separator. Insulate the tank all around, unless you can bury it fully or at least part way. If you can bury it to some extent, then you might take some advantage of the soil heat from below by insulating just below grade, and out about 4’ horiz. all around the tank. Do you have any plant ‘waste heat’ that you could pipe through the fluid in the tank? Can you just up the heat on the inlet pipe line, since that equipment already exists? All of this will take some experimentation on site and depending upon the ambient temps.

 

[bimr]

Electric heat trace or steam will be necessary as well as foam glass insulation. Foam glass insulation is necessary to avoid wicking up water into the insulation. Insulate all sides. Install a removable cover on the top

 

[BridgeSmith]

If you have, or could make available, hot water, you could do something similar to residential hot water radiant heat systems. You embed pipes in to the tank walls and/or floor, or even within the tank itself, if feasible, to heat it with hot water or steam. That way you don't have to worry about trying maintain electric heating elements exposed to water.

 

[oldestguy]

I'd check out the problems with insulation and advantages of some. In any case insulation can be the main way to keep things stable, in spite of power losses, etc.

 

[bimr]

If you have, or could make available, hot water, you could do something similar to residential hot water radiant heat systems. You embed pipes in to the tank walls and/or floor, or even within the tank itself, if feasible, to heat it with hot water or steam. That way you don't have to worry about trying maintain electric heating elements exposed to water.

Have never seen this done, probably because it is too expensive.

Self limiting heat tape is done all of the time on the exterior of tanks and pipelines.

Electric heat tracing

 

[BridgeSmith]

"Have never seen this done, probably because it is too expensive."

Possibly not for tanks like this one, but it is done sometimes for in-floor heating systems in residential construction. It's not overly expensive to construct and fairly economical to operate.

"Self limiting heat tape is done all of the time on the exterior of tanks and pipelines."

Of course it works fine on the outside, but not very efficient without alot of added insulation. The thermal mass and heat transfer characteristics of the concrete make heating the outside much less efficient than heating the water directly. The best system depends on the priorities for the system, whether it's a low initial cost, enhanced durability, low cost operation, ease of operation and maintenence, etc. I wasn't necessarily advocating for a particular system, just throwing out an option for consideration.

 

[bimr]

If there is enough heat present in the wastewater to keep the oil water separator from freezing, you can get by with just foam glass insulation/ with weatherproof insulation cover as well as a cover on the top on the ows.

If the wastewater will freeze, then you need to consider steam tracing or electric tracing in addition to the insulation.


From the above comments, it would appear that you have never seen an ows. An ows looks something like what is pictured below:

Steam or electric tracing is installed on the outside of the tank as the wastewater fluid will foul anything submerged in the wastewater.

 

[BridgeSmith]

"From the above comments, it would appear that you have never seen an ows."

You got me there; no I am not familiar with how they are configured. However, there was this: "The tank may be cast-in-place concrete, so heat tracing the exterior might be inefficient." in atrizzy's second post. I therefore assumed that embedding steam or hot water lines in the concrete would be an option. I didn't know what conditions or mechanisms would be inside the tank, which is why I qualified my suggestion of placing pipes in the tank with the phrase "if feasible".

 

[atrizzy]

bimr, we're proposing a passive API style gravity-functioning OWS. I have seen and designed many out of cast-in-place concrete.

 

[bimr]

Sorry, but I missed that.

This reminds me of being brought late into a Russia wastewater project where the Indian based consultant had designed an elevated concrete ows. Not very feasible trying to construct such a design. The Indian water pollution market is a generation behind in technology because of a lack of funding.

Most environmental designers ended the use of ows's constructed of concrete back in the 1950's as concrete is permeable and the oil will seep through and contaminate the concrete and ground. Coatings will not attach well to concrete either. Industry doesn't like concrete either as it is too slow to construct and is expensive. Finally, if made of concrete, you would have to use an expensive liner like the anchor-lok system.

anchor-lok-lining-system

The concrete with a liner would make a nice installation, but will be expensive. But then, insulating and heat tracing a concrete tank is not practical as concrete doesn't have good thermodynamic properties. You would also have the concern that the ows is sitting on the ground and you will never know if the concrete is leaking through the bottom.

Very few ows are now modeled after the traditional API design. Most industries have abandoned building these large gravity settling vessels because they are simply too costly to justify. Traditional API units have an excessively large footprint and are far too much of a real estate burden to consider as a viable option. The large API design units are also considered to be a fire hazard by the insurance underwriters and would have to be located away from other process units. There will also be an air emissions issue and a cover would be necessary for that reason.

API-Oil-Water-Separators-Simple-But-Costly

I would strongly suggest that you recommend to your client a modern ows installed on a flat concrete pad, off the ground and w/secondary containment.

 

[LittleInch]

Yes, I would go with what bimr is saying.

I haven't seen a concrete trough type API separator for decades and in my early days was responsible for replacing many failed units with the tilting plate single metal construction as bimr showed in the previous posts.

It would be feasible to insulate those and provide a reasonable level of heating to prevent freezing.

I still can't really understand how in minus 40C you're actually going to get any water flowing willingly into these units which are typically for low oil quantities in surface water.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[atrizzy]

Duly noted. The site I'm looking at currently uses several old concrete OWS's. I'll look into proposing something more active and modern.