Purely activated carbon as a rechargeable water recycler
Activated carbon – and other adsorbents- can remove almost any contaminant from water. To some degree.
The problem is that for some substances the amount it can remove before the carbon is “full” is impractically low. To remove any substantially practical amount would take way too much carbon, if you were just going to throw the carbon away when it got “full”.
Salts of any sort are one of these things. Salts are pretty much the worst group of compounds, with regards to how much can be removed by adsorption. However, the thing about physical adsorption is that it is usually a completely *reversible* process.
If you were to run perfectly clean water backwards though the activated carbon bed, the contaminants come right back out again! The stream of water coming out would have pretty close to the same concentration of contaminants in it as what the greywater had.
So obviously this is pretty useless in a completely self contained water treatment system, where you had no source of clean water (if you did you would use that instead of trying to clean the greywater!), because it takes just as much water to clean the carbon as you got out of the water recycler. The thing is, *the contaminants are stored in the carbon* in the meantime, in a much smaller space and with less weight than it would take to store the contaminated water.
So what if you used this instead of holding tanks for an RV – you run the greywater though a big carbon filter, getting the water back so you can reuse it, no holding tanks needed. When the carbon gets full, then when you get somewhere with tapwater available, run water backwards through the system to clean out the activated carbon, making it usable again!
The question is how much carbon you would need to clean up the greywater to a degree where it can be reused. If it took a gram of carbon for every gram of water cleaned there would be not much point – the system would be weigh just as much as holding tanks.
However, I have been looking into this, and while trying to work out the details, it looks like you might be able to get roughly a ten to one ratio or maybe a bit less i.e. 10 kg of carbon would just be like carrying around 100 Kg of water. Not bad, clearly there are some uses for that. Especially because carbon is fairly cheap.
There are a few hitches, however. One is that tap water is not perfectly clean. That seems to be the biggest one. That could produce undesired effects, and reduce the usable level of this exact system -carbon only- to below a ten to one ratio, which I’ll explain in a sec.
The way I came up with the 10 to one figure is this:
The limiting contaminant in this case is the salts, ignore the organic contaminants, detergents, whatever, because they are easily adsorbed and won’t be what determines the size the carbon filter must be.
For each adsorbate (contaminant in the water) and adsorbent (in this case activated carbon derived from cashew shells and made in a specific way) pair you can draw a graph called an isotherm, which tells you, at a given temperature and PH, and if there were no other complicating factors involved, which you can think of for the time being as a rough approximation of how much of the contaminant can be adsorbed per gram of carbon in a carbon filter vs. the concentration of the substance in the solution entering the activated carbon (AC) bed.
As I said, salts, or rather ions in solution, are the worst case scenario. Obviously with any salt there are at least 2 ions when it is dissolved, but that is taken care of actually, so just ignore that for a sec.
There are many complicating factors, but if you were to look at all the different isotherms for all the different contaminants in the water, and consider the concentration of each contaminant in the water, and compute the amount of carbon you would need to treat a gram of water according to each isotherm, and then use the lowest amount as the limiting factor, to decide how much carbon you need to treat a gram of your water, multiply that by how much water you want to treat, and that is a way to get a very rough estimate of how big your carbon filter has to be.
We know ions are the most poorly adsorbed, so if we just assume for a second that cadmium’s isotherm is the worst case scenario, I checked the numbers for cadmium using numbers from this document. I just used Cd because it is easy to get the isotherm for. I tried to get the one for potassium, which is even worse adsorbed, but couldn’t find it. But K isn’t toxic or anything anyway so it’s no big deal if it is left in in modest quantities.
Suppose the greywater had about 1 mg/gram of salt in it (salt is all ions when dissolved in water), and all the ions are adsorbed at least as well as cadmium, then using the constants given in that doc and the equation at the wikipedia page on the freundlich isotherm, the carbon bed can store 0.095 mg/g of ions. With 1 mg/g of ions in the water, that’s enough to store the contaminants contained in 10.5 grams of water in each gram of carbon. So about a ten to one ratio.
The thing is, this will only go down as the quantity and variety and quantity of other substances goes up, and they compete for adsorption sites. However it shows that it might be worth doing a small scale test (and it is generally acknowledged in industry that a small scale test is usually the only practical way to find this out):
Just take a small carbon block filter, run water slowly through it until the water coming out the other end is no longer satisfactorily clean ( maybe measure by evaporating samples of the water and seeing what is left behind). Take the amount of water that had gone through filter and divide by the mass of carbon.
Then try regenerating it with a slightly larger volume of tap water.
At the end of the day, if adsorbing every contaminants including salts does not work well enough, you could use mixed ion exchange resin to remove them and still do the backflush- carbon thing for all other contaminants, making a water recycler that consisted of nothing except a big carbon filter and some ion exchange resin. Which would be quite cheap and of course easy to build.