Out With The Herms And In With The Rims

[jjeffrey]

I know a lot of people love their HERMS setup, but I loathed mine. I have a 2 vessle setup- gas fired kettle and mash/lauter tun. I use the kettle also as a HLT. I had an immersion coil which sat in the kettle, through which i'd pump my wort. Kettle would heat/hold sparge water whilst I switch the pump on or off to regulate temperature in the mash.

It worked ok, but I found it cumbersome for a number of reasons:

1. Manual handling. My kettle is elevated, and I needed to lift an 18m long copper immersion coil filled with wort in and out, often with the kettle hot. There's no way I would have got away with that sort of operation at work, it wasn't safe- WorkSafe would be on my arse in a flash. When the coil wasn't in the kettle I had to find space on the floor for it, and I was constantly tripping over it. The long hose length required also created clutter. I know some people would (and have) say "why don't you just set up a separate HLT and keep the coil in there instead of creating clutter on the floor?". Well, in doing that, I'm not really reducing my footprint, and it increases setup complexity (something I'd prefer to avoid).
2. Cleaning. More hose, more coil, more cleaning. I hate this aspect of brewing.
3. Water heating time. Heat your sparge water and control the mash at the same time thus saving time, right? It takes me a little under an hour to heat 60L of sparge water, so the times don't work out for me. This means I generally gotta heat up 80L BEFORE I dough in. Usually I'm still twiddling my thumbs come 10AM.
4. Temperature response in the mash tun. Unless I got the water in the kettle to boiling (i.e. too hot for sparge), my ramp times were way too long.

I started off just trying to make the HERMS more efficient, but the more I thought about it, the more I realised that a RIMS would pretty much solve everything. So I made this:





Specs:

• Element: 2.5kW single phase low density element, 600 long, 25mm in diameter. It's pretty big but with low heat output so that I can minimise the surface temperature (and hence maillard reactions) whilst still getting the heat into the fluid. I based the size on a couple of degrees every 10 minutes, whilst still being able to suck enough juice out of a standard power point to do the job. This was the most expensive bit of the whole kit, representing 70% of the overall cost.
• Control via PID controller with high speed solid state pulse output. Temp probe is in the end of the unit (RTD).
• Cleaning real easy. the whole thing comes apart by undoing the triclovers.

Does it work? Don't know. Just after I took these photos I went to start circulating but found that I'd put one of the check balls back to front in the diaphragm pump when I cleaned it last- now I've gotta pull it apart and fix it before I start. I'm glad I went to try with water first- it would have been bad had I 12kg of grain in the tun whilst I mucked around with the pump. Hopefully I'll have it commissioned later today.

jj.

 

[Stubbie]

Nice piece of engineering there, JJ.

I find myself stuck deciding whether to go HERMS or RIMS, though my present leaning is towards RIMS. So I'm interested to hear how yours performs.

I'm looking at elec element options atm. Can you provide some more info on your element - eg. model and supplier?

Stubbie.

 

[MHB]

Nice

I'm guessing that's a food grade air powered diaphragm pump, if the element was the dearest part, you must have scored well to get that for less than 4-500 bucks.

BTW where did you get the element? I'm looking for something similar.

MHB

 

[jjeffrey]

MHB: yeah, the wilden pump was a good score, although I've been using it for yonks and it isn't included in the overall cost of the RIMS setup. I got it for free from a wool scouring plant that was shutting down in Geelong years and years ago (the method by which I score most of my gear). It was previously used as a hydrogen peroxide dosing pump, although it did cost me $250 to kit it out with the food-grade+high temp seals, balls and diaphragms.

The element was custom built by Hotco in Dandenong- type in Hotco to google and you'll find them. Basically I told them what to build and they made it for me. The specs are:

• 2.5kW
• low density coil (6W/cm2)
• 1" ODx600 long S.S. sheath
• 1" BSP brass end adapter sealed into a Clipsal junction box. The adapter I screwed into a 2"x1"BSP Triclover Benny Socket to suit the rig.

It's a custom which I opted for because from an engineering perspective I'm anal and I had the budget alotted. I was particular concerned with high surface temps as my pump is a little slow as well as cleaning, but plenty of other people have MUCH cheaper/hotter standard elements which they swear by. Thirstyboy is one of these. The use of a solid state relay and high pulse control will also alleviate some of these problems (not to mention be heaps better for your element).

 

[MHB]

Cheers i've been talking to Stokes but I will give them a call to.
The old rule of thumb for wort boiling was 1W/inch^2 of element to avoid scorching or about 0.16W/cm^2. I think in a pumped system with turbulent flow (looks like that's what you have designed in) you should be able to run far higher heat densities, especially as you are well below the boiling point, be good to avoid having the surface of the element so hot you get micro boiling (easy with a good controller) if you hear that characteristic rumbling noise you will know to take the foot off the gas.

Nice work and I would hijack that pump in a heart beat

Mark

 

[Thirsty Boy]

Nice unit mate - very pretty.

I get away with the re-purposed kettle element... but that thing is ideal. I'm still long term planning to re-do mine with an LD element.... couple of other things higher up the list though.

We had a pump like that lying around work for ages - unused and neglected - after 18 months I was just starting to look at it with avaricial eyes when someone tidied the damn thing up (probably because they are worth so damn much money) - should have been faster!!

You brewery is really starting to come together proper like now.

Cheers

TB

 

[pbrosnan]

The old rule of thumb for wort boiling was 1W/inch^2 of element to avoid scorching or about 0.16W/cm^2.[/font]

It's a funny thing but you hear of wort scorching every now and then but it's a bit hard to find any examples of it actually occurring. What is the formula you're quoting based on?

 

[jjeffrey]

It's a funny thing but you hear of wort scorching every now and then but it's a bit hard to find any examples of it actually occurring. What is the formula you're quoting based on?

Definately. When I was designing this thing, I could not get a single first hand testimonial about wort scorching with RIMS units (I did, however, get lots of hearsay, not uncommon for this craft). In the end I decided to spend the entire alotted budget on best practice, although what really helped my decision in going with an element like this was a simple, easily cleaned cartridge element. I just picked the lowest commonly available heat density.

I didn't quote any formulae, but my calculations for the element were based on this procedure:

1. How much juice can I get out of a single phase general power outlet? The answer is about 10A (based on the earth pin rating), although my breakers trip at 16A. Using Ohms law, P=IV=10Ax240V=2400W=2.4kW. I bumped it up (slightly), and chose 2.5kW.
2. What does 2.5kW give me in terms of heating power? Use standard heat capacity equation, q=mCpDt, where q is energy kJ, m is mass of stuff to heat in kg, Cp is heat capacity- i assumed that for water which is 4.18 kJ/(kg.C), and Dt is temperature change in C. if you choose a time basis of 10 minutes (or 600 seconds), q=600*2.5=1500kJ. you can then use the heat capacity equation to calculate how much temp change you get for a given mass in your mash tun. Transposing you have Dt=1500/(m*4.18). for 30 L of liquid mash (ie, 30kg), Dt=1500(/(30*4.18)= 12C in 10 minutes. For 60kg (my capacity), it works out to about 6C. Am I happy with this? You betcha.
3. How big does the element have to be? Choose a diameter- I chose 1"OD based on ID of tube I could use for the RIMS casing, in which the element would be housed. From the Hotco site, the lowest heat density available was 6W/cm2. Ignoring the ends, the surface area of the element is A=2.54*pi*L in cm2, where L is length in cm. Also, A=P/heat density=2500W/6(w/cm2)=417cm2. Therefore the length I need was 417/(2.54*pi)=52cm, or 520mm if you don't want to be a dress maker. If you allow about 80mm at the lead end of the element so that the mounting thread doesn't get hot, you are looking at 600mm long.

That's what the element sizing was based on.

jj.

 

[jjeffrey]

Phase 1 commissioing complete. Once I reassembled by pump properly and got flow, I put 20L of cold water into the mash tun, punched in 65C and let her rip. I got 1.5C per minute. The controller needed to be re-tuned (it was tuned for the heat straps on my fermenter), but afterwards I only got 0.1C overshoot for a +5C step change, and then a flat line bang on setpoint for rest of the testing time (10 minutes).

happy as. next trick will be to see if I can burn some oatmeal stout sweet mash liquor onto the surface of the element.

jj.

PS- PBROSNAN, it occured to me after pushing the POST button on my last post that you were querying MHB rather than me. The "EDIT" button is not where I remember it, so i left it (plus, I reckon heat transfer calcs rock, which was why I enjoyed writing it in the 1st place).

 

[jjeffrey]

Well, this thing has been sitting in my garage for a month and I FINALLY got to use it on something other than water (this month has really killed me).

With the 3rd day of the boxing day test periodically blurting over the air compressor, I tried to gum up my new element on a 60L batch of oatmeal stout.

From an operational point of view, the unit was a breeze. I did forget to turn the heating circuit on till about 10 minutes after striking, at which point I was reading 63 (i struck a little low). But she came up to 68 within a couple of minutes, then stayed there. Mashed out at 78. Took about 6-10 minutes to get there from 68.

So, the big question, did it scorch? Have a look at the photos. These were taken before any rinsing was done.



What you can see there is a thin film of stout extract. I gave it a squirt with the hose, and it was like I had never used it. I got zero scorching. The bloody thing was still shiney after a rinse (no cloth used).

On Friday I'll be testing it on a 4 step Bohemian pilsener. Hopefully by then the volcano in my fridge has settled down a bit.

jj.

 

[Bizier]

I reckon heat transfer calcs rock

That leaves me feeling both in awe and inherently stupid. Good thread jjeffrey.

 

[Thirsty Boy]

That there is loverly. The difference between compromising a little in order to save money, and actually buying the best thing for the job.

One of these days.....

TB

 

[absinthe]

It's a funny thing but you hear of wort scorching every now and then but it's a bit hard to find any examples of it actually occurring. What is the formula you're quoting based on?

i dont think the wort will scorch unless you have a stuck mash when you are not watching.. the temp would drop on the probe and so the element will stay on with some captive mash around the element and that could cause problems

 

[jjeffrey]

i dont think the wort will scorch unless you have a stuck mash when you are not watching.. the temp would drop on the probe and so the element will stay on with some captive mash around the element and that could cause problems

Yeah, I was worried about that. The trick is to install an interlock sensing element (either flow or temperature) as close to the element as possible, such that if the flow stops or the temperature gets high in the vicinity of the element, it cuts the juice. For my setup, I did this by installing the temp probe about 20mm from the hottest part of the element (the end), and installing the body of the unit at an angle so that in the event of no flow, the hottest liquor will rise to the probe. The angle also allows any air to come out without priming. The probe is installed on the underside to ensure it's always immersed. My controller has an alarm contact which cuts power to the element if the temperature exceeds 2C above set-point.

I tested that function on this batch- I got a total of 5C overshoot when I shut the feed valve before the control loop reacted. The process took about 5 seconds.

Ironically, this was the first time I've brewed this recipe and NOT got a stuck mash- usually a combination of wheat and oats plays havoc on the grain bed. And I decided not to do a b-glucan rest too cause the last time I got a drop in mouthfeel as a result. Hail LOTS of rice hulls.

jj.

 

[jjeffrey]

That there is loverly. The difference between compromising a little in order to save money, and actually buying the best thing for the job.

Cheers. That's awesome, coming from a bloke who has a rims with 300lb ANSI SS flanges on it. Why don't you come round on Friday whilst the moravian is mashing, and I'll let you look at it

By the way, I got a filter. I WILL brew a better beer than you

One of these days.....

jj.

 

[roller997]

That looks like an excellent result.

I think you may have swayed some HERMS supporters over to a RIMS setup.

It looks like the element was worth every cent and with your setup a stuck mash is no longer a problem either from a scorching perspective which is great.

Great thread - Thanks for the information.

Roller997

 

[jjeffrey]

Pilsener batch completed today- RIMS performance was very interesting.

It was a Bohemian with Weyermann Pilsner malt and Carafoam. about 15kg of grain with 40L of water. The aim was to do a protein rest at 50C for 15 minutes (the European pilseners has stuck on me in the past), ramp it to 62C for 20 minutes, then 67C for another 20 minutes, then to mash out at 78C.

I was 1C over on strike, so the RIMS didn't really do much to begin with except hold temp. Then I punched in 62 and off she went. It got to 55C in about 4 minutes, but then seemed to slow right down. It took a further 15minutes to get to 62, so I abandon the b-amalase step and punched in 67. In all, it took 30 minutes to get to 67C, but once there was was able to hold ok. The mash-out ramp was slow too: after 20 minutes it was at 76C, which is when I called it quits. In the kettle I noticed exceptionally large protein flocs. The colour came out as a nice straw at the end of boil.

On cleaning, I found this:





I was like a coating of tripe over black scorched areas. Rinsing the "tripe" off, you can see what sort of scorching I had:




And here's a shot after a run-in with the curly girl:


So, I did get scorching, but it didn't seem to affect the colour (in this instance), just the performance. Time will tell what it's done to the flavour. Interesting that I didn't get that with the stout. The conclusions I draw from that are that you don't get trouble from sugars, but you do from proteins+sugars. ALSO, cartridge elements are way easy to clean, AND make sure your RIMS casing is as easy to clean as the element (mine was- took me a minute to clean it thankfully).

jj

 

[Thirsty Boy]

exactly what I get - Fraser John also posted a thread a while ago and it sounded like this was happening in his rims too. Normally, it is a really light layer of the tripe stuff (which I believe is simply break material forming on the surface) but as you noticed, when it builds up, it insulates the element, impacts heat exchange performance and at the extremes burns a little onto the element.

For me - it was far worse when I changed from the braid to the false bottom, and FJ seemed to get more trouble when he had a turbid wort with a high flour content. So I think wort turbidity might have something to do with it. Maybe just giving the proteins particles to form around??

Next time I brew its my plan to keep the RIMS element off during my strike and initial circulation - so I will mash in at 55C as per usual, re-circ to set my grain bed and let the wort clarify a bit.. then RIMS on to ramp to first sach rest. That suits my mash schedule because I want my P rest to be a short 3-5mins anyway - then ramp. No real time for temp loss, but plenty to get things settled down and running clear.

Easy technique tweak - instead of going to the PID and punching in 67... I do that and flick the power supply switch while I'm there.

I'll talk to you about how it goes.

TB

PS - the worst its ever been.. still couldn't taste it in the beer. Or I couldn't anyway, maybe because I don't want too??

 

[jjeffrey]

TB, do you get it with all malts, or just certain types? I agree it's break formation and thus some malts are going to be worse than others. Obviously the mash regime will also effect it because of element control.

I've some theories why my stout batch didn't scorch:
a] The element didn't really work hard until mash-out because it was a single infusion mash where the rims just pulsed to maintain temp (the pulses are just a fraction of a second). Perhaps any break that formed wouldn't stick on because only small amounts were being generated at any one time, and it would have been washed into the grain bed. Perhaps by the time I mashed out, there wasn't enough break material left to cause grief when the element is on all the time during the ramp. Certainly I couldn't see much break in the kettle, but my stout is blacker than black, so it wouldn't be obvious.

b] the Joe White Pale malt that I used for the bulk of the grain bill in the stout was relatively low in soluable protein (must check specs against the Weyermann Pilsener- to be confirmed)

c] In the pilsener batch, there was a clear point at which the performance seized up, and it was around 55C (but the film temperature of the element would have been way higher). Also it was about 4 minutes after the element was on full blast (no pulsing). I wonder what temperature the break forms. It may be that if you strike at a higher temp, the protein coagulates in the grain bed and gets trapped rather than forming on the surface of the element. Perhaps the low temp rest exacerbates this because the degradation of protein and b-glucan matrices in the endosperm may release more soluable or shorter chain proteins into the wort.

My next few beers will be single step pales. I'll go at a variety of temps and use different base malts over a series of batches to see how the RIMS runs. I'm also going to install a pulse timer, so that the element can't be on full blast for more than a certain period of time (say 30 seconds or so).

jj

 

[jjeffrey]

Weyermann Bohemian Pilsener vs. Joe White traditional ale looks very similar in terms of protein content, although slightly (almost marginally) less modified.

ref:View attachment Malt_Specs.xls

This leads me to believe that the difference in scorching with the past 2 batches was more process based rather than malt based.

jj.