Determining Conversion Efficiency in "All-in-one" electric systems

Homebrew Talk - Beer, Wine, Mead, & Cider Brewing Discussion Forum

Help Support Homebrew Talk - Beer, Wine, Mead, & Cider Brewing Discussion Forum:

Noob_Brewer

Well-Known Member
Joined
Mar 4, 2019
Messages
1,392
Reaction score
3,423
Location
Eastern, NC
The Question: when is the most appropriate time to take a refractometer reading to calculate conversion efficiency in all-in-one systems, before or after you lift the malt-pipe?

I am a new brewer having started in january and I have the anvil foundry 10.5g system with recirculation package. I thought at first to post this in one of the foundry threads but thought that since the general designs are similar across all-in-one systems like this, it might reach a broader audience here. After reading countless times by @doug293cz helping people trouble shoot their efficiencies, I started measuring and calculating conversion/lauter efficiencies myself for each batch while beer smith is used to calculate mash/lauter efficiencies.

So heres what I do: 1) I calculate the theoretical Fmax using braukeiser's formula. I then monitor the progress of the mash, refractometer readings, several times throughout the mash duration. I take my final reading right before I lift the malt-pipe and I take another reading after the wort has drained mostly (but before any sparging). I have been calculating the First wort extract (braukeiser terminology) AFTER I lift the malt-pipe but before I sparge to then compare this BRIX reading to the FMAX to calculate conversion efficiency (Fmeasured/Fmax*100). The problem with this with this system is that once you pull the malt-pipe you leave some sugars behind of course and thus the BRIX reading drops slightly. See the attached table I created with the last 4 consecutive brews. You will notice that in beer "C" the brim reading dropped quite a bit. This has happened to me with this drastic of a change 3-4 times before as well. So I started thinking that not only are sugars being left in the grain after pulling the malt-pipe but not all of the water is being mixed with the wort during the mash because of the gap between the malt-pipe and the kettle. I confirmed this issue before by taking a refractometer reading taking the liquid between the malt-pipe and kettle. Its virtually sugarless!

So while taking the BRIX reading AFTER lifting the malt-pipe is suboptimal then taking it before lifting should be better? Well again, if some of the volume of strike water is NOT mixing with the wort inside the malt-pipe then the wort is essentially a thicker water/grist ratio. Again, if you look at "Beer C" in the table below - you will notice that the refractometer reading at the end of the mash before I lifted the malt-pipe was actually higher than the theoretical Fmax. SO this told me that there was likely a lower volume of water mixing with the grains than the total volume of strike water used in the mash and in the calculation of Fmax.

If you look at my mash efficiencies and the general grists that I have been using from the table, I am very happy with the outcomes thus far because mash efficiencies of ~75-78% is pretty good in my book. The reason why Ive wanted to calculate conversion/lauter efficiencies is to help optimize my setup though: and I enjoy being a nerd and taking lots of measurements! :)

In anticipation of these questions: 1) I do trust my refractometer readings. It is calibrated and validated against my hydrometer readings which I always still take pre-boil and into-the-fermenter. 2) I am super meticulous about achieving my target volumes preboil, post-boil, and into the fermenter. I have made a calibrated measuring stick to help me determine strike volumes and any volumes etc. So I don't think this issue is with erratic differences in volumes (actual vs measured).

So any thoughts on the best way to to estimate First Wort Extract in the calculation of conversion efficiency would be welcome. I have thought of modeling how much water might not be mixing with the wort to reduce the water used in the volume part of Fmax. However, I don't think this is uniform across brews.

Thanks in advance! Also thanks @doug293cz for your thoughtful posts on these issues and posting out braukaisers work!

EDIT: I forgot to add that I have settled on a gap setting of 0.028 using a cereal killer mill and I do use a wilder bag in the malt pipesince I do have a good bit of flower. I also recirculate after the first 10minutes but VERY slowly and I do this only to help maintain mash temps.

1592433192426.png
 
Last edited:

doug293cz

BIABer, Beer Math Nerd, ePanel Designer, Pilot
Staff member
Mod
HBT Supporter
Joined
May 14, 2014
Messages
13,639
Reaction score
10,652
Location
Renton
The Question: when is the most appropriate time to take a refractometer reading to calculate conversion efficiency in all-in-one systems, before or after you lift the malt-pipe?

I am a new brewer having started in january and I have the anvil foundry 10.5g system with recirculation package. I thought at first to post this in one of the foundry threads but thought that since the general designs are similar across all-in-one systems like this, it might reach a broader audience here. After reading countless times by @doug293cz helping people trouble shoot their efficiencies, I started measuring and calculating conversion/lauter efficiencies myself for each batch while beer smith is used to calculate mash/lauter efficiencies.

So heres what I do: 1) I calculate the theoretical Fmax using braukeiser's formula. I then monitor the progress of the mash, refractometer readings, several times throughout the mash duration. I take my final reading right before I lift the malt-pipe and I take another reading after the wort has drained mostly (but before any sparging). I have been calculating the First wort extract (braukeiser terminology) AFTER I lift the malt-pipe but before I sparge to then compare this BRIX reading to the FMAX to calculate conversion efficiency (Fmeasured/Fmax*100). The problem with this with this system is that once you pull the malt-pipe you leave some sugars behind of course and thus the BRIX reading drops slightly. See the attached table I created with the last 4 consecutive brews. You will notice that in beer "C" the brim reading dropped quite a bit. This has happened to me with this drastic of a change 3-4 times before as well. So I started thinking that not only are sugars being left in the grain after pulling the malt-pipe but not all of the water is being mixed with the wort during the mash because of the gap between the malt-pipe and the kettle. I confirmed this issue before by taking a refractometer reading taking the liquid between the malt-pipe and kettle. Its virtually sugarless!

So while taking the BRIX reading AFTER lifting the malt-pipe is suboptimal then taking it before lifting should be better? Well again, if some of the volume of strike water is NOT mixing with the wort inside the malt-pipe then the wort is essentially a thicker water/grist ratio. Again, if you look at "Beer C" in the table below - you will notice that the refractometer reading at the end of the mash before I lifted the malt-pipe was actually higher than the theoretical Fmax. SO this told me that there was likely a lower volume of water mixing with the grains than the total volume of strike water used in the mash and in the calculation of Fmax.

If you look at my mash efficiencies and the general grists that I have been using from the table, I am very happy with the outcomes thus far because mash efficiencies of ~75-78% is pretty good in my book. The reason why Ive wanted to calculate conversion/lauter efficiencies is to help optimize my setup though: and I enjoy being a nerd and taking lots of measurements! :)

In anticipation of these questions: 1) I do trust my refractometer readings. It is calibrated and validated against my hydrometer readings which I always still take pre-boil and into-the-fermenter. 2) I am super meticulous about achieving my target volumes preboil, post-boil, and into the fermenter. I have made a calibrated measuring stick to help me determine strike volumes and any volumes etc. So I don't think this issue is with erratic differences in volumes (actual vs measured).

So any thoughts on the best way to to estimate First Wort Extract in the calculation of conversion efficiency would be welcome. I have thought of modeling how much water might not be mixing with the wort to reduce the water used in the volume part of Fmax. However, I don't think this is uniform across brews.

Thanks in advance! Also thanks @doug293cz for your thoughtful posts on these issues and posting out braukaisers work!

EDIT: I forgot to add that I have settled on a gap setting of 0.028 using a cereal killer mill and I do use a wilder bag in the malt pipesince I do have a good bit of flower. I also recirculate after the first 10minutes but VERY slowly and I do this only to help maintain mash temps.

View attachment 685429
Excellent analysis. One of the principle requirements for Braukaiser's method for conversion efficiency to be accurate is to have homogeneity of the wort throughout the entire volume when the SG samples are taken. In systems with a solid sided malt pipe, this is difficult to achieve, even with recirculation. In a recirculating mash system without a pipe, homogenization is pretty much a given. For a simple BIAB (no false bottom, no recirculation, bag fills entire vessel volume) aggressive stirring before taking a sample is sufficient. In an MLT with false bottom, and no recirculation, stirring will not be adequate, but vorlaufing about 4X the volume under the FB should do a pretty good job of homogenization. I haven't thought about the best way to achieve homogenization in the case of a solid sided pipe.

Brew on :mug:
 
OP
OP
N

Noob_Brewer

Well-Known Member
Joined
Mar 4, 2019
Messages
1,392
Reaction score
3,423
Location
Eastern, NC
for Braukaiser's method for conversion efficiency to be accurate is to have homogeneity of the wort throughout the entire volume
Well damn! I was hoping for some sort of lightning rod answer with infinite wisdom! lmao :) Anyways, based on my refractometer readings from between the malt-pipe and kettle, it is obvious the entire volume is NOT homogenous when the malt-pipe is in the kettle. Again, I thought of modeling, with some gross assumptions, just how much mash water isn't mixing with the rest of the wort BUT 1) I don't think this is consistent across batches and 2) when i took those refractometer readings on about 3-4 batches, the BRIX readings wasn't zero but was just VERY low. My best estimate is that 1-1.5g of water is above the mesh sides of the malt-pipe where the water is not mixing (or at least VERY diluted) with the wort. However, when you lift the malt pipe to let drain, I recirculate the wort within the kettle (not dumping it back into the malt-pipe but directly into the kettle) to make it homogenous as possible AND the whole volume is there (subtracting water absorbed into the grain). So my question now is: is the magnitude of sugars left in the grain after raising the malt-pipe something I could account for if it is generally constant OR at least correlated with grain bill size? if this is the case I think that I could better estimate the conversion efficiency after lifting the malt-pipe if I 1) subtract the water lost due to grain absorption and 2) apply a correction factor for the sugars lost. Just thinking out loud here so I could be way off base.

Thanks for your response though!
 

CBurgles

Active Member
Joined
Jun 6, 2018
Messages
31
Reaction score
1
Location
Hyde park
I have the M&B and have had the same conundrum. If I take a gravity reading before pulling up the malt pipe it is somewhere between 7-10 points higher than after I lift the malt pipe. I’ve tried taking sample from the top of the mash, the middle of the grain bed, straight from the spigot, off the recirculation arm, and all give the same result. Gets to be a little frustrating when you check and think you’re at or over your target so you decide to end the mash, but turns out you’re still low.
 
OP
OP
N

Noob_Brewer

Well-Known Member
Joined
Mar 4, 2019
Messages
1,392
Reaction score
3,423
Location
Eastern, NC
I am beginning to wonder if the best way to tackle this issue with all-in-one systems is to estimate a proportion of wort at gravity "x" (inside and below the malt pipe) vs gravity "y" (outside the malt pipe - but perhaps only the top half of the height of the malt-pipe), and combine the two to arrive at the predicted gravity (with relatively more weight to the gravity inside the malt-pipe.). So gravity measurements at multiple locations to predict the actual gravity before pulling the malt-pipe up.

For example, the last beer I brewed, had a BRIX measurement of 15.8 taken from the recirculation arm at 65minutes, but the BRIX reading from the top of liquid level between the malt-pipe vs kettle wall was 8.0 on the nose. So obviously, there IS mixing, but it is much diluted.

Thinking out loud on this...hmmm...
 

doug293cz

BIABer, Beer Math Nerd, ePanel Designer, Pilot
Staff member
Mod
HBT Supporter
Joined
May 14, 2014
Messages
13,639
Reaction score
10,652
Location
Renton
Would it be possible to put an asymmetrical "Y" in the recirc return, such that ~90% of the wort flows into the malt pipe, and ~10% flows into the gap between the pipe and kettle wall? This would continuously mix the wort in the two zones during the mash. Setting up the gap flow so that the incoming wort flows tangentially to kettle wall would probably be optimal.

I don't have one of these units so can't test this idea myself.

Brew on :mug:
 
OP
OP
N

Noob_Brewer

Well-Known Member
Joined
Mar 4, 2019
Messages
1,392
Reaction score
3,423
Location
Eastern, NC
Would it be possible to put an asymmetrical "Y" in the recirc return, such that ~90% of the wort flows into the malt pipe, and ~10% flows into the gap between the pipe and kettle wall? This would continuously mix the wort in the two zones during the mash. Setting up the gap flow so that the incoming wort flows tangentially to kettle wall would probably be optimal.

I don't have one of these units so can't test this idea myself.

Brew on :mug:
Interesting idea. I suppose you could do this. So this approach would focus on trying to better homogenize the gravity of ALL of the water/wort rather than trying to simply account for the differential in gravities among two different "zones". hmmmm, will think on this idea, although its getting more complicated lol
 

doug293cz

BIABer, Beer Math Nerd, ePanel Designer, Pilot
Staff member
Mod
HBT Supporter
Joined
May 14, 2014
Messages
13,639
Reaction score
10,652
Location
Renton
Interesting idea. I suppose you could do this. So this approach would focus on trying to better homogenize the gravity of ALL of the water/wort rather than trying to simply account for the differential in gravities among two different "zones". hmmmm, will think on this idea, although its getting more complicated lol
Exactly.

Brew on :mug:
 

doug293cz

BIABer, Beer Math Nerd, ePanel Designer, Pilot
Staff member
Mod
HBT Supporter
Joined
May 14, 2014
Messages
13,639
Reaction score
10,652
Location
Renton
After thinking some more about the non-homogeneity issue with malt pipes, something that might work to homogenize the wort prior to SG sampling would be to raise the pipe, let it drain well, and then lower the pipe back into the wort. A few cycles of this should be pretty effective at mixing up the wort.

Brew on :mug:
 

kevin58

Supporting Member
HBT Supporter
Joined
Sep 4, 2017
Messages
1,360
Reaction score
1,017
@Oginme tackled the issue of efficiency and how the water between a malt pipe and system side wall do not mix using a standard recirculation in the Anvil Foundry. The solution is simply to lift the malt pipe up once in the middle of the mash and once just before mash-out. He reports an efficiency in the mid to high 80% range when doing so.
 

Fhizzicks

Member
Joined
Jul 7, 2020
Messages
9
Reaction score
11
Location
Oakland
I imagine the solution of lifting the malt pipe part way through the mash to better homogenize the wort would also boost overall efficiency as you won't be over saturating the little water that is actively recirculating. I'll have to try that next time on my Grainfather. I saw a drop in efficiency on a few brews where I slowed down my recirculating rate; now I may know why.
 

doug293cz

BIABer, Beer Math Nerd, ePanel Designer, Pilot
Staff member
Mod
HBT Supporter
Joined
May 14, 2014
Messages
13,639
Reaction score
10,652
Location
Renton
I imagine the solution of lifting the malt pipe part way through the mash to better homogenize the wort would also boost overall efficiency as you won't be over saturating the little water that is actively recirculating. I'll have to try that next time on my Grainfather. I saw a drop in efficiency on a few brews where I slowed down my recirculating rate; now I may know why.
Saturation of wort with sugar is not an issue. At mash temps, the saturation point for maltose in water is 67°Plato (something north of 1.300 SG.) There is no way you can get the SG that high in a mash. With a mash thickness of 1 qt/lb, the highest SG you can get in the mash is ~27°P or ~1.115. That's less than half way to saturation.

Brew on :mug:
 
OP
OP
N

Noob_Brewer

Well-Known Member
Joined
Mar 4, 2019
Messages
1,392
Reaction score
3,423
Location
Eastern, NC
After thinking some more about the non-homogeneity issue with malt pipes, something that might work to homogenize the wort prior to SG sampling would be to raise the pipe, let it drain well, and then lower the pipe back into the wort. A few cycles of this should be pretty effective at mixing up the wort.
@Oginme tackled the issue of efficiency and how the water between a malt pipe and system side wall do not mix using a standard recirculation in the Anvil Foundry. The solution is simply to lift the malt pipe up once in the middle of the mash and once just before mash-out. He reports an efficiency in the mid to high 80% range when doing so.

So yes, I thought of this myself. But to be honest, when I was using the bag in malt-pipe, I had the bag edges over the malt-pipe and kettle rim and it just seemed to me to be more work and potential for "accidents" (even with a proper hoist) with the 10.5gal foundry to do this just to mix the water around the outer edges with the rest of the wort. So this just didn't appeal to me personally. My original intent with this thread to was to try to rectify how to mix the water that wasn't mixing with the wort to get a better idea for conversion efficiency measurements, based on Braukeiser's formulas, really with any AIO systems because they all have similar issues with the wort not being homogenous. I have since changed how I brew with the anvil foundry with ditching the malt-pipe all together and simply using the bag in the kettle with a brewzilla false bottom. Very happy with this setup - the wort is homogenous as it can be now and my conversion efficiencies are routinely near 100% and as a result my mash lauter efficiencies are in the low 80s consistently. Best of all, without the malt-pipe, you can have more grains for higher OG beers. I brewed a imperial stout (just under 22lbs of grains) some time ago with this setup, sparged a couple gallons and hit 80% mash/lauter efficiency on the nose. So very pleased with my current setup with out using the malt-pipe.
 

tracer bullet

Supporting Member
HBT Supporter
Joined
Aug 10, 2020
Messages
1,663
Reaction score
1,439
Location
Minnesota
I brewed a imperial stout (just under 22lbs of grains) some time ago with this setup, sparged a couple gallons and hit 80% mash/lauter efficiency on the nose. So very pleased with my current setup with out using the malt-pipe.
@Noob_Brewer :) I've brewed a ton on my 6.5, I now have a 10.5 but haven't used it yet. I'd like to pick your brain so to speak and know what to expect if I go for a 22 lb grain bill.

I'm guessing you use a bag and a false bottom. Or maybe not the false bottom. Do you get a decent water to grain ratio, where it settles and has a little water on top / not oatmeal? Do you have much room to stir, or is it a bit scary to keep it from spilling?
 
OP
OP
N

Noob_Brewer

Well-Known Member
Joined
Mar 4, 2019
Messages
1,392
Reaction score
3,423
Location
Eastern, NC
@Noob_Brewer :) I've brewed a ton on my 6.5, I now have a 10.5 but haven't used it yet. I'd like to pick your brain so to speak and know what to expect if I go for a 22 lb grain bill.

I'm guessing you use a bag and a false bottom. Or maybe not the false bottom. Do you get a decent water to grain ratio, where it settles and has a little water on top / not oatmeal? Do you have much room to stir, or is it a bit scary to keep it from spilling?
Yes, I use a bag (wilser) and the brewzilla false bottom which allows for recirculation while maintaining temps quite well. The stout I was referring to above was my first stout, my second stout I used 24lbs of grain + 1lb rice hulls = 25lbs total in 7.69gallons of water so that water to grist ratio was ~ 1.23qt/pound. So it was quite thick AND took up a tons of space.

First pick is right after mashing in. Tight! after recirculating for a while it settled down a bit. I sparged with 2.5gallons of water to another vessel which allows me to sparge with the bag in place and then lift the bag once sparging is done. 8gallons pre-boil. funny enough with your timing, when I brewed this one I said to myself, thats the limit! Tomorrow I'm brewing another stout and going with 24.75lbs of grain lol.

IMG_0288.JPG

IMG_0290.jpg
IMG_0294.jpg


IMG_0295 (1).jpg
 

tracer bullet

Supporting Member
HBT Supporter
Joined
Aug 10, 2020
Messages
1,663
Reaction score
1,439
Location
Minnesota
Thank you so much! This is perfect.

I've got a 14 gallon cooler (mash tun) I have had a long time, and planned to keep, but then realized it would actually just be needed for a RIS 1-2 times a year. That made me wonder, realistically, how much grain would comfortably fit the Anvil (to check the realistic answer vs. the "well I did it and set a record but it was terrible" answer). Looks like some pretty high ABV's can be had with a simple-ish approach (mash & sparge).

OK, thanks! Cooler not needed and can probably be sold w/ the propane burner someday. I'll wait until I make an RIS in the Anvil but it really helps to know what I can aim for.

Thanks again -
 
OP
OP
N

Noob_Brewer

Well-Known Member
Joined
Mar 4, 2019
Messages
1,392
Reaction score
3,423
Location
Eastern, NC
Anytime. FWIW I think anvil recommends a 16lb grain limit when using the malt pipe but when I was using a bag inside the malt pipe I regularly pushed it to 19-20lbs (including rice hulls) lol. Without the malt pipe you get a bunch more room and the water that was on the sides of the maltpipe is now included and more homogenous with the mash. With my brewzilla false bottom I have about a gallon below it as it sits on top/just above the outlet valve after modifying it
 

tracer bullet

Supporting Member
HBT Supporter
Joined
Aug 10, 2020
Messages
1,663
Reaction score
1,439
Location
Minnesota
Anytime. FWIW I think anvil recommends a 16lb grain limit when using the malt pipe but when I was using a bag inside the malt pipe I regularly pushed it to 19-20lbs (including rice hulls) lol. Without the malt pipe you get a bunch more room and the water that was on the sides of the maltpipe is now included and more homogenous with the mash. With my brewzilla false bottom I have about a gallon below it as it sits on top/just above the outlet valve after modifying it
Yeah, I do get all that. For the 6.5, and I'm sure the 10.5 is identical (same diameter just taller):

1 gallon of space under the malt pipe
30% of the rest (the cross sectional area the body of the pipe inhabits) is actually outside of the malt pipe

I use Excel to do my calculations, a sheet I developed long ago and just keep tweaking. I have multiple ratios that get calculated - grain vs. water overall, grain vs. water if the grains are contained in the pipe, and similar for if just the bag is used (for me meaning a bag with a false bottom). For a RIS for sure it'll be bag, no pipe. That's why I was happy that you included your water volume as well!

If you're curious I (well, my spreadsheet) calculated 1.23 qts / lb overall, 1.07 in the bag area alone if a false bottom is used, and 0.76 if the pipe was used (it wouldn't actually fit in the pipe, so that's sort of irrelevant). And that about 9.94 gallons of space would be taken up overall which seems fairly accurate judging by your picture.
 
Last edited:

Brewdog80

Well-Known Member
Joined
Sep 17, 2021
Messages
236
Reaction score
180
My question is just why? If you are wanting a big brew, just drop from 5g batch down to 4. 20% reduction in grain, so 20 instead of 25, or even drop to 3 gallon and you can make it bigger. And not run the risk of a major mishap and fry the electonics
 

Brewdog80

Well-Known Member
Joined
Sep 17, 2021
Messages
236
Reaction score
180
Ha, it's like using rice hulls to try and prevent stuck mash. Malt, even when ground fine, is heavier than rice hull which ain't doing anything to help....
 
OP
OP
N

Noob_Brewer

Well-Known Member
Joined
Mar 4, 2019
Messages
1,392
Reaction score
3,423
Location
Eastern, NC
My question is just why? If you are wanting a big brew, just drop from 5g batch down to 4. 20% reduction in grain, so 20 instead of 25, or even drop to 3 gallon and you can make it bigger. And not run the risk of a major mishap and fry the electonics
Although I agree with @tracer bullet on enjoying pushing the limits, for me its also this: I only brew these BIG stouts ~1x per year and I want them to last me the year as I taste them throughout. So while you are correct in that I could get a bigger beer by going to 4 gallons, I simply want a full 5 gallon keg if Im only doing this 1x per year. With that being said, it is truly amazing to me that with some modifications, the 10.5gallon foundry is quite capable to delivering a 5gallon finished batch of beer with an OG of ~1.102.
 

tracer bullet

Supporting Member
HBT Supporter
Joined
Aug 10, 2020
Messages
1,663
Reaction score
1,439
Location
Minnesota
Ha, it's like using rice hulls to try and prevent stuck mash. Malt, even when ground fine, is heavier than rice hull which ain't doing anything to help....
It's not ball bearings in applesauce. The malt doesn't sink. You mix the hulls and grains before they go in, and while you put them in, and they stay mixed. It's quite effective even if you don't see why.
 

Brewdog80

Well-Known Member
Joined
Sep 17, 2021
Messages
236
Reaction score
180
Nope, if you mix, they float, malt sinks....waste of money and space. And I'm fine with only brewing big once in a while, but I don't have room in the two keg fridge for a big one to sit all year so I'd have to bottle, and it ain't happening....
 

DuncB

Well-Known Member
Joined
Jun 5, 2019
Messages
2,673
Reaction score
1,542
Location
Paremata New Zealand
Would it be possible to put an asymmetrical "Y" in the recirc return, such that ~90% of the wort flows into the malt pipe, and ~10% flows into the gap between the pipe and kettle wall? This would continuously mix the wort in the two zones during the mash. Setting up the gap flow so that the incoming wort flows tangentially to kettle wall would probably be optimal.

I don't have one of these units so can't test this idea myself.

Brew on :mug:
I have done this on my Guten all in one. Have a T on the recirculation pipe and the side arm goes to the whirlpool which mixes that "deadspace " water between the malt pipe and vessel wall.
This definitely improved my efficiency, Brew your own had a feature on improving efficiency in brew systems. , I'd already done the modifications before their article. But my mods are in the july issue.
 

aceluby

Supporting Member
HBT Supporter
Joined
May 24, 2022
Messages
165
Reaction score
249
Well, that's not my experience. A few may make it to the top at first but they get water logged and sink and stay quite well mixed.

Your mileage may vary as they say.
No mileage about it. It’s how physics works, He might not find value in it, which is fine, but what he’s describing physically can not happen unless he’s throwing stuff in after mash in and never stirring.
 
Top