EcoRenovator (
-   Conservation (
-   -   sewer water heat exchanger (

oil pan 4 02-28-16 07:58 PM

sewer water heat exchanger
The average temperature of household waste water is typically between 80F and 90F.
All you have to do to figure out how many BTUs you are flushing down the toilet is to get a water bill figure out how many pounds of water you used, subtract ground water temperature from average waste water temperature. Then multiply the temperature differential by pounds of water.
I bet it comes out to be a pretty big number.
What if we could put some of that heat back into the household?
If some one were to live off grid this could massively reduce their water heating requirements and save a lot of propane.

Here is my idea:
Get a good long length of 4 inch, 0.080'' wall 304 stainless steel food grade pipe, because most home poo pipes are 4''. Then take a length of 2.5" or 3" and split it down the middle and weld it on to the bottom of the 4 inch pipe, then weld water line hook ups to the ends of this bulge. So as the warm waste water flows out cool fresh water comes in the waste gives up its heat to fresh water.
I have about 100psi of water pressure so I want 80 wall pipe, if you had well water at a much lower pressure then standard 1/16'' pipe would be fine.
Did I mention that I may try to build this?

To build this you need lots of 4 and 2.5 or 3 inch pipe.
I have been told this 4 inch 80 wall pipe runs upwards of $20 per foot retail prices and 2 or 3 feet isn't going to be enough, this is a heat exchanger so it has to be long, as long as possible.
Lucky for me I can get a lot of pipe and cheap as dirt from the scrap yard. I already have about 40 feet of food grade stainless 2.5 inch pipe I have some 0.080'' wall 4 inch pipe but its slated for another project, I should be able to get a lot more.

I don't have to use 0.080 wall pipe but I kind of want to.

Other hazards to look out for:
When you weld stainless in the presence of oxygen you can create hexavalent chromium if you don't know what you are doing. If you don't know already what hex chrome is then you should not be trying this.
The welds have to be air tight.

mejunkhound 02-28-16 10:23 PM

I bet it comes out to be a pretty big number.

Hmm, more like pretty small. Run the numbers.

oil pan 4 02-28-16 10:54 PM

According to the EPA the average American family uses 400 gallons per day.
That is 99,600 pounds of water per month.
Heat that up 30'F for a total of 2,988,000 BTUs.
Or the energy of almost 1 gallon of gas per day, every day for years.

Seems like a pretty decent sized number to me.

mejunkhound 02-29-16 12:37 AM

I'd probably do what you are doing if I had 'free' stainless steel (or copper) pipe and disregarded the salvage value.

Using your epa numbers, , you are leaving out an awful lot of factors.

70% used indoors, so that drops the 400 to 280 gal

Leaks and toilet are not heated water, epa says those are 40%, leaves 120gal day

epa says clothes washers are 22%, will assume warm water left at end, I'll grant your 30 F there, but we wash in cold water. 1/2 of faucet taken as cold water, 8%, leaves 120-32 = 88 gal per day that may be warm enough, say 100F (70+30) to recover heat.
Even engineered heat exchanges for liquid to air need about 10F differential,

so, max you get is 20F * 88 gal *8.3# = 14,608 BTU per day, about a pint of fuel, or say 45 cents a day if you have an elec WH, less if nat gas heater. A long long time for ROI, if ever, just based on scrap value of SS pipe.
Take away the clothes washer or use a modern one, the number drops to near zero since the water going down the drain from a shower is not sure 100F unless you have alligator skin.

oil pan 4 02-29-16 01:24 PM

There are other appliances like the dish washer that use hot water, most only have 1 water connection and its recommended it be connected to hot water. Then if you connect them to cold water then they just use their built in heating element to heat the water.
Toilet water can be heated in some cases to prevent moisture from condensing on the toilet running down the toilet causing mold or wood rot.
For me I get oil, grease, hydraulic fluid, diesel on clothes so the real dirty stuff gets washed in the hottest water possible.

I already bought the 316 alloy 2.5 inch tube last year for 70 cents a pound.
When the scrap 4 inch pipe becomes available it should only be around 50 cents per pound because its only 304 alloy. 8 to 10 inches of 0.080 wall 4 inch pipe is about 1 pound. If I cant get enough 0.080 wall then I will use 0.062 wall, which should be closer to 1lb per foot.

I am thinking this is going to be a bigger energy saver for winter time.
Ground water comes into the house any where between 40'F and near freezing, even our natural gas hot water heater cant keep up during winter.
The summer time water temperature differential is only around 30'F, but during winter it will be nearly double that.

gtojohn 03-01-16 02:46 AM

What would you heat with the waste water?

jeff5may 03-01-16 07:44 AM

The most efficient implementation of this strategy is to preheat incoming potable water supply on its way to the water heater. To account for time lag (as with clothes or dish washing), it is more effective to wrap a "worm" tube around the drain pipe section. The longer the length of time the water spends in the worm, the better the usage vs discharge lag is dealt with.

The other common method is to preheat the cold line of a sink or shower with the drain water at its point of use. There have been many builds documented where the shower floor was embedded with hydronic tubing, warming the incoming water with the shower drain water. Some were constructed the same as heating slabs,and some had intimate, direct contact between the drain water and the incoming tubing. Both methods worked well enough to notice the preheating effect.

oil pan 4 03-01-16 12:18 PM

I don't see any point in wrapping a worm around the sewer pipe. All the out going water in the sewer pipe is almost never deeper than 1 inch and does not stay in the pipe for very long unless you have problems.
If a worm were wrapped around the pipe only the edges of the 2 pipes would be in contact, through the walls of the pipes.
The way I propose is more like a heat exchanger.


Originally Posted by gtojohn (Post 49366)
What would you heat with the waste water?

Incoming utility water.

oil pan 4 03-05-16 02:18 AM

5 Attachment(s)
Time for a visual aid.
Here are some small cross sectional mockups I did.
Looks like 2 inch pipe on 4 inch pipe wont work. Its going to have to be 2.5 inch pipe.
I used actual 0.080 wall 4 inch pipe and standard 0.062 wall 2 and 2.5 inch pipe.

2.5 inch pipe on 4 inch pipe:

2 inch pipe on 4 inch pipe:

Some welds. Miller maxstar, 70 amps, 2% Thoriated tungsten, very pointy, 20SCFM.
To further improve the welds I am thinking about getting plasma cutter torch wheels and putting them on the tig torch that way I can roll the torch along the weld and get a super smooth evenly heated weld almost as if a laser wielding robot welded it.

Next step will be to build about a 1 foot long functional prototype, that wont be put in the ground, just torcher tested to failure.
Where I will do actual back purging the fresh water pipe and develop a good way to split 2.5 inch pipe down the middle and fit up the welding joint a little tighter.
Then I can measure the volume of the 1 foot section of fresh water pipe and see how much of a hot water reservoir I have.

stevehull 03-05-16 06:37 AM

Oil pan,

Hate to be a wet blanket, but your return in $, heat or whatever is going to be negligible. Up north, where entering cold water is COLD, then heat recovery from waste water can be rational.

I built one for my shower when living in Michigan about 30 years ago. The cold water coming in was about 40 F and the drain water in the shower was about 80 F. I plumbed it so that the recovered heat, preheated the cold water to the shower valve. After about ten minutes in the shower, I noticed I could turn down the hot water valve just a tiny bit.

It worked - but not to my expectations.

The key is understanding the difference in water temperatures. Assume your incoming water temp is about 60 F. Assume that waste water is about 80F. The difference that is there (delta T) to recover is about 20F or so. The recovery is no where near 100% so let's assume 50%. That means that a 10 F difference is there. Now multiply by the average number of quarts per minute going down the drain when this is working - let's call it two gallons (8 quarts) per minute. The number of BTUs saved is tiny! About 80 BTUs per minute.

As an example, the heat from one burning match is about 1 BTU.

Now go to a more northern place where incoming water temp is about 40 F (all else same). The delta T is now 40 F - double that in your area. The recovery is also double, but still small.

I love the idea of recovering heat, but sometimes the cost (or % recovery) to do so is just not there. Waste sewer water might only be 60 or 50 F.

But as an interesting and fun project? There I am with you 100%!!

You will need to mount your recovery tube in a vertical orientation so the water swirls down on the pipe surface (water "skin effect"). Copper will be a far better heat transfer media compared to other materials. Even better would be to make a coaxial heat exchanger with ripples or dimples on the exchange metal to expand the available surface area for heat exchange. A long pipe (10 feet) would also help to get efficiency. But count on a maximum of maybe 50%.


oil pan 4 03-05-16 09:51 AM

Yeah if I build this thing it's going to be fairly large. I have 60 feet of sewer pipe. To be worth anything I know it will have to be at least 20 feet long.
Only 2 things limit the size :
The physical length of 60 feet.
The availability of 50 cent a pound scrap stainless.

I expect efficiency to only be between 20% and 60%.

Even if all it does is 10% efficiency and prevent the pipes from freezing abs bursting then it will have been worth it.

I don't think that copper would be strong enough. All industrial heat exchanges are made of stainless.

oil pan 4 03-28-16 08:27 PM

I picked up 30 to 40 feet of 0.083'' wall stainless today from the scrap yard.
I did not get very much 2.5 inch pipe and I am pretty sure the 2.5 inch pipe I got is not 80 wall.
So there isn't really any reason why I cant. I thought the main hang up was going to be finding enough 80 wall 4 inch pipe.

stevehull 03-28-16 09:06 PM

Before you go to all this trouble, PLEASE verify your waste water heat value. You claim it is between 80 and 90 F.

Waste water recovery works best when the water inlet temp is cold, preferably below 45 F. I believe your water inlet temp is about 65 F.

Do not get the peak temp of waste water, get a 24 hour average. Not hard to do. THEN, design the heat recovery system.


oil pan 4 03-29-16 01:03 AM

Here the city some how has gotten away with burring the water line less than 2 feet.
So during the winter we get water that is ice cold. Its not freezing but close. The natural gas water heater acts more like an electric, as in it cant keep up.

A 24 hour average is kind of useless since water is not flowing out 24 hours a day. I could run each house hold device, pop the top off one of the sewer clean outs and point my flir at the out going water.
My wife takes raging hot showers, that one would be the most interesting to see going out.

jeff5may 03-29-16 10:14 AM

The "piggyback" hx you mocked up will do fine on horizontal drain runs. It should live 100 years or more as a drain line. Just make sure the fresh water channel has a way to burp itself. Your heat transfer surface lies directly where air will accumulate if allowed to.

In an industrial hx, they use stainless because it is stronger and less expensive than copper. It is not an awesome heat conductor, so when the unit has to be compact, they use cupronickel (copper stainless) or (you guessed it) copper for the working transfer tubing. In modern industry, less expense translates to more return nearly always.

stevehull 03-29-16 03:17 PM

All you need to do (to measure the outgoing sewer water temperature) is to buy one of the inexpensive digital temperature recording devices (see below), put it in a plastic freezer bag and weight it down in the sewer outlet with a string attached. Then retrieve it in a few days.

It will show you the maximum, minimum and average temperature of the water over the data collection period.

Here is a simple RC-5 recorder that works well for me.

RC 5 USB Temperature Data Logger Temp Recorder Internal Sensor 32000 Points New | eBay

When you speak of "freezing cold" winter water - is this a temperature or a subjective assessment? For example, if I get into the shower and it seems "freezing cold" it might be 80 F.

Metrics make all the difference in planning . . . .


oil pan 4 03-30-16 03:55 PM

1 Attachment(s)
I also picked up a high frequency starter/stabilizer box.
This will allow me to light the arc with out touching the base metal with the tungsten electrode.
Now I also have a gas saver kit for my WP-17 torch.

The idea is I will use the gas saver and a strip or angle iron held in place with a vice grip to guide the torch along weld, running hot, fast and perfectly straight creating a weld that looks like it was done with some kind of automated laser machine.
I want the welds I put on the actual pipe I am going to use to look a lot better than my quick and dirty test.

To split all the 2.5 inch pipe I think I will use a cut off wheel. It will be slow and will suck a lot but that will give me straight cuts with out warping the pipe. I already know if I use my plasma cutter it will bow the pipe. I may test filling the pipe with kitty litter and cutting it but I still expect the pipe to bow.

I am thinking my first live prototype will be a shower water heat exchanger. it has been a while since I have been up under the house, and if I remember correctly the shower drain pipe is 2 inch pvc goes for about 3 feet before it goes into the main line. The cold water line is also very close by.
It would be easy to make heat exchanger section and plumb it in there.

mejunkhound 03-31-16 11:49 AM

Keep track of how much time you spend on this, I'd be curious.

NOte: have done some projects myself that I figured my 'return' was about 12 cents per hour - but it was 'fun'. <G>

oil pan 4 04-02-16 12:57 AM

The thing with this project is once it's done it should work night and day for decades.
So far the only time I spent on this was building the mockups in the pictures, maybe 2 hours while I was at work one night when I had to stay late because no one on night shift bothered to show up to work. As long as I keep moving and do things the are visually stimulating no one bothers to ask what I am doing.
So in a strange way the time I spent on this project far has pretty much paid for most of the materials I need to build it for real.

All times are GMT -5. The time now is 12:43 PM.

Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2021, vBulletin Solutions Inc.
Ad Management by RedTyger