I recently have received several inquiries around the sizing of resin columns and amount of resin necessary for a particular flow rate that ensures proper resin function over the life of the media - I therefore thought it was probably worth its own topic or discussion:

The initial thing to keep in mind is biodiesel contact time with the ion exchange media that being said resin weight, density, volume and expansion become important...

Rohm & Haas was the original North American player and they quote numbers of ~50 lbs per cubic foot which equates to the typical gel resin density of 800 g/liter (6.66 lb/gal) however individual batches of resin come with some variance and a typical range is 800 to 830 g/liter - I usually use 815 g/liter as my base number which yields 6.92 lb/gal.

Also there is some question as to determining the proper flow rate per pound or bed volumes (BV) particularly because of the swelling factor in resin and how much they do or don't swell during biodiesel purification...(please see the following some what involved explanation of what I mean by this)

You should notice that in the Rohm & Haas literature they quote flow rates both ways: the 0.36 gal/pound of resin and they say starting flow rates of 2 BV/hr. This actually equates to 0.30 gal/pound of resin.

Purolite which also sells a fairly equivalent gel resin for biodiesel purification states it this way in their brochure:

How much Purolite® PD 206 do I need?
A Purolite® PD206 polishing system is sized based on design flow for processing biodiesel. The minimum size recommended is a system that will handle 3 bed volumes per hour (BV/h). Since glycerin is absorbed slowly, flows of 1.5-2 BV/h may be advised. (Caveat: Consult your Purolite representative for complete details.)

Here you might also notice they use 3 Bed Volumes per hour as design criteria but then turn around and recommend using 1.5 to 2 BV/hr flow rates (for a safe margin in glycerin adsorbtion).

We have an exclusive North American distributorship agreement with Thermax Corporation and work very closely with them and we have discussed this in some detail: the question of flow rate is some what curious since biodiesel purification and water purification are not the same thing at all.

With water purification you could count on the gel resins to expand (swell) to the maximum capacity of almost 300 percent. So flow rates on resin were easy to determine since the resin whether dry or expanded always performed the same but biodiesel isn't responsible for the resin's expansion instead it is the methanol. And eventually it is the deposited glycerin or adsorbed glycerin which determines the size of the resin beds overall expansion.

This being the case the resin in the presence of a gross amount of methanol will swell to ~105 perhaps even 110% of its original size but during biodiesel purification typical methanol percentages range from 1% to maybe 4%. And then as it adsorbs glycerin this displaces methanol and reduces the overall size of the beads expansion. So as the resin does its job there can actually be bed size contraction if it reached its fuller expansion early on.

Bed Volume and operating flow rate are ultimately important since it will determine the average residence time (or contact time) the biodiesel has with the ion exchange sites on the resin beads.

So since methanol content and subsequently a resin bed's expanded volume can vary greatly I have chosen to always use the initial unused bed volume to determine the optimum operating flow rate.

That all being said and going with a conservative average gel resin density of 815 g/liter yields 6.92 lbs per gallon of gel resin media and I use 3 times this dry weight knowing that the actual bed volume will expand to nearly twice this volume which by defacto yields the more conservative (recommended) flow rate suggested by Purolite of 1.5 BV/hr.

If you think about it, this makes a lot of sense since early on in the resin beds life, when at its smallest bed volume (unexpanded), it will adsorb glycerin the most readily since it is empty so to speak. Then as the resin bed swells with use and glycerin adsorbtion, the effective BV/hr flow rate slows naturally with the resins decreased capacity for glycerine adsorbtion. It's self correcting in this way...

GCG

Hi,

Firstly, really good and informative article. Im going to do a water wash after my reaction 2 through my biodiesel to absorp the residual methanol in the biodiesel. Then separate it.

If I do one water wash before the dry wash stage, will this significantly improve my resin life?

Would I be able to get away with just one resin exchange tower and if not happy with product quality just recirculate it through the resin tower?

Thanks a lot, its an expensive investment so I want to get it right the first time.

Cheers,

Reece,

We just traded a ion exchange tower to a friend of ours that has always done water washing and is very experienced at it - he also does a 2 stage or (80/20) type reaction process and then water washes and then dries.

After getting one of our dry wash towers he also thought that he would rather just do a small water wash and then process through the ion exchange towers to finish it off. He didn't want to use the 5% prewash because he recovers methanol from the seperated glycerin and he doesn't want to degrade the purity of the methanol recovered by adding water to it from a pre-wash.

Anyway he found that dropping out the glycerin and then doing a small wash cylce still left a fair amount of water in the crude biodiesel and this water is not good for an ion exchange resin purification process since the resin will have a pretty hi affinity for the water and therefore the water can become the most limiting factor in your purification process.

Having said that though methanol and water our 100% miscible and will both saturate the ion exchange resin so what will happen is you will initially make good fuel but shortly the biodiesel will begin to have moisture content and will probably needed additional drying after ion exchange purification.

Sunbreak Biofuels out of Oregon actually has about a 1% water content in its fuel prior to ion exchange purification and the T-45 BD macroporous resin only achieved about 65% of its normal loading prior to requiring flushing...

Gel resin in this particular case would probably be preferrable since it expands more and has a higher retention capacity for the water...

GCG

Thanks GCG,

Yeah I will be using Purolite PD 206 as they mention how it acts as a desiccant (which is the main reason I want it - don't want to use air dry methods as getting our biodiesel to ASTM oxidative stability limits has been the only test we fail).
In a biodiesel trailer set up I went to (they give seminars) they do a one day process and then pump the biodiesel into their truck to drive to their next location. Pretty cool.
They used purolite to dry it quickly, and that gave me the idea.

Thanks again for your detailed reply, I appreciate your help.

Reece

Reece,

What you need to do, if you want to determine the effectiveness of using gel resin as a desiccant - simply figure out how many grams of water can be adsorbed per lb of resin and then test your fuel for its parts per million (equivalent to mg/kg or for simplicity a mg/liter of biodiesel).

Please also realize that glycerin will displace water that has been adsorbed and reintroduce it to the biodiesel...

So if 1lb of resin expands 100% with water then it will hold about 1.15 lbs of water (based on simple specific gravity differences between a gel resin and water) or 522 grams.

This is equivalent to 522,000 mg so if your biodiesel contains 5000 ppm (.5%) of water then 1 lb of gel resin can dry 100 liters of biodiesel.

Keep in mind this doesn't account for the 100% miscibility of water and methanol nor does it account for the glycerin being adsorbed in the same 100 liters of biodiesel.

GCG