Help choosing and building the right ethanol set up

CannaBusiness
, , , ,
1

So I posted on the forum a week or so ago to field interest in split processing for crude. As I expected, I got an underwhelming amount of responses from people that were interested. I’m sure we all understand the position of most people in the industry given the collapse since last year and I would prefer not to harp on that factor too much. The lack of viable options has lead me to believe I should simply build my own set up.

I have roughly 60k lbs of biomass myself and my good friend has another 40k of his own. Obviously there is also no shortage of bio supply. I was thinking it would be good to have a set up that could do 500-1000lbs a day into crude. I could utilize this for myself and maybe some friends as well. Does anyone have any recommendations as far as well priced effective equipment goes? Maybe a used set up that they have out grown or are planning on replacing? I am open to all recommendations and information that anyone is willing to provide.

Given that I am ultimately undereducated in the process of building a lab, I would also like to hire someone to help me streamline the process of turning this idea into a reality. I have learned more than once that it is best to do this proper the first time around because it costs much more to fix a mistake after the fact.

Thanks everyone for always sharing your gems and helping new folks get started. I have learned a lot from this forum just staying in the background and am grateful for you all.

2 Likes
2

I’d go centerfuge n ffe. Message me if you need help I’m in ky.

3

Is your end product crude oil or winterized crude oil? That would be the first step in defining your process.

If you are going to end at winterized crude, you can extract with ethanol at low temperatures, winterize post extraction or use membrane filtration to remove fats/waxes. If you are ending at crude oil, extract with room temperature ethanol and save some money.

I would agree that centrifuges are a good option for your target capacity, but do not expect the biomass to be dry after extraction (expect at least 20% by mass loss of ethanol to biomass). From centrifuge then particulate filttation, then solvent recovery (typcially in a falling film bulk solvent recovery system), desolvating (typically done under vacuum and higher heat). I work at Aptiaengineering.com but there are several equipment providers who would tell you similar process flows.

If you would like to talk further, I would be happy to help: info@aptiaengineering.com
Good luck

1 Like
4

If this is really the case then all of those centrifuges people are spending 70K on are a waste of money. Ya’ll are doing this wrong.

When I clean out my fuge, the biomass is powder dry. 95%+ recovery rate.

6 Likes
5

Yeah. I get 80% recovery even without using a centrifuge, just drip drying. That’s an absurd statement

4 Likes
6

Well if you’re trying to extract hemp like it’s a sack of dirty clothes in the washing machine i’m not surprised by anything.

$70k for a system like that is a joke.

My fuge does it differently and really does get things drier. yields are great too.

1 Like
7

I’m familiar with that method of centrifuging; it definitely works well and is definitely my preferred implementation. But even a soak and spin should do way better than 80%.

2 Likes
8

I’m not trying to say anything bad about centerfuges: I’ve never used one. But I lose 2-4% etoh when draining our ~50-80lb reactor (using chilled etoh). If I’ve done pulls correctly, this is really like 1%. Vacuum pulls, air pressure pushes, and gravity doesn’t really leave much to lose except that the biomass is a sponge itself. That’s where our loss happens, and I think it’s low enough to live with until we become larger.

9

The major benefit of a good fuge over this is the time saved. Turning up gravity is easier than pushing and pulling and pushing again.

Hard to beat spinning at 2000+ rpms

1 Like
10

I think you should look at your numbers again.

I agree this is where your losses are happening. I don’t believe you’ve done the math correctly. how much ethanol is going in to that reactor? (80gal?) how much is coming out? (65-70gal free flow would be my guess). when I was extracting in kegs, even a three day drip dry was leaving 15% or more of my starting solvent in the biomass. using vacuum to recover the solvent doesn’t get the cannabinoids that were in said solvent.

Are you running that solvent over another batch of biomass? if not, you’re recovering 3x more solvent than necessary. if so, you need to get your spent biomass potency tested because you’re likely leaving a good chunk of the goodies behind.

11

The math is not really right, but I’m not losing above 10%. I’d get fired. My pulls and pushes pull most out if I am following procedure. But we definitely lose some, I’m sure future system revisions as we build will correct this loss to be more palatable.

12

I’m curious. How long does it take you to do an entire run?

13

Hey @driftlessoil ,
The centrifuges are excellent at removing most of the solvent but I do stand by the averages that I have seen. I have weighed biomass in and out of a centrifuge to determine these numbers.
Example: 15 kg biomass load going in - 18 kg biomass & solvent coming out. That’s 3 kg solvent or more left in biomass after you consider that oils have been removed. 3 kg of ethanol is ~ 1 gallon.
Sure there is some variability in these numbers depending on grind, material quality, spin time etc. I would be interested to learn how you tested your residual solvent if you are willing to share. What RPM and G force are you generating? I see from your later post that you are getting 2000 + RPMs but without your Gs it is hard to compare? What is your spin time?

I would expect the recovery in a centrifuge to far exceed a drip drying method in both recovery rate (obviously) and actual amount recovered (evaporation is hard to account for if you leave it drip drying for a long period of time and then take a post drip dry mass). @Magisterchemist if you would explain how you are drip drying I would be interested to learn your testing methodology and drip tech.
In general, I tend to offer a more conservative numbers when communicating numbers that people will use to build their economic models because people sometimes come in optimistic.

The other discussion is extraction efficiency which is a level of how much CBD is left in biomass after extraction. How much CBD are ya’ll leaving behind? There can be some variability in this one also but on average, what are you testing at in your spent biomass?

Hope everyone is staying safe and well.

14

Do you know how much G force or rpm & diameter of your fuge?

15

We commonly use a centrifuge which spins up to 1,600 RPM and creates a max G force of 715. I don’t know the basket diameter off the top of my head. This is a 15 kg biomass per run centrifuge. So this centrifuge does spin more slowly than the previously mentioned 2000 RPM option driftless was discussing.

16

That might be your issue right there. I think you want about double that G force to get things dryer. What’s your dry cycle time? I usually found 4-5 minutes on full spin got me down to diminishing returns on my spin. It’s a smaller basket at 2800rpm though

1 Like
17

I respect your conservative numbers. At the very least, you’re compensating for losses elsewhere in the system. People in this space are often overly optimistic.

If the averages you are looking at come from typical centrifuges designed like the delta cup, then I believe you. IMO those systems are flawed to the core. And spending 70K on a system that is flawed is the wrong move. Even if it is shiny and marketed directly to us.

We are extracting plant material, and trying to avoid undesirables. Our target compounds are on the outside of the cells and the undesirables are on the inside. Why would you want to “clean” your hemp like a sack of dirty clothes? When you flood and soak and agitate, all you are doing is forcing solvent into the biomass where all of the undesirables are.

I use a shop vac to empty it, and haven’t taken the time to weigh the spent biomass. I will note that it is legitimately powder dry when finished however. It is still flammable with a slight smell of alcohol so it is not 100% recovery, but it is far from wet.

I’m lacking funding, so I also haven’t tested my spent biomass. :upside_down_face:

I do measure the ethanol in and the tincture out. I should be a bit more analytical here, but honestly I have bigger fish to fry. 7 gallons in 6.5+gallons out. I’m pretty sure most of my “loss” is in the pipes and on the floor or evaporated, not in the biomass. I guess that is 92%, so maybe I lied above :lying_face:

Like I said, I think the typical centrifuge usage is wrong. So instead of flooding, agitating, and trying to spin dry a giant sponge, I spin my biomass and then flow solvent over it in recirculating manner. That way there is continual change of solvent over the biomass and with the increase in gravity, it doesn’t have much of a chance to soak it up. The solvent stays on the outside of the cells this way. Most of my undesirables come from the fines after milling. I wish I could do something different here.

I am using 10-11% CBD biomass roughly milled
10 Kg biomass per load
5 gallon wash -70C
2 gallon rinse -70C
total wash/rinse time is under 5 minutes
I combine 2 runs to make one batch so here are some real numbers from a recent batch.
21.22 KG biomass in
2.125 Kg crude oil out - (2.125/21.22) *100 = 10.01% yield by wt
COA tells me 68% total CBD (65% CBDa) and 73.5% total Cannabinoids

I spin dry at top speed for 15 minutes, but solvent stops coming off at 12 minutes.
There’s math here to tell me how fast I’m spinning and the G-force created, but I don’t have it. Probably over 1000 Gs?
There’s also tools that would tell me, but I don’t have them either. I wish I did.
Also, since things aren’t changing direction and spin the whole time, my motor doesn’t work all that hard with momentum in its favor. Some simple mods that are above my head would allow this system to spin even faster, recovering even more. Where it stands now it works great, but it could improve even more.

1 Like