RIMS Build

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

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

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

Sizz

Well-Known Member
Joined
Jan 10, 2008
Messages
137
Reaction score
2
Location
Kennesaw, GA
Update 09-10-10

The system is complete aside from a few aesthetic and convenience modifications! This thread is fairly long so I'll summarize the key points here.

This is an all electric fly-sparge based two-vessel system utilizing a MLT & kettle. A high wattage 5500w electric element is used as a dual purpose external heater. The first role of the heater is for heating the recirculated mash. The 2nd role is to directly heat sparge water from ground temps to mash out temps.

Key advantages of the system:

Smaller footprint since there's only 2 vessels.
Quick mash stepping.
Energy efficient, since water is heated directly as needed.
Drain is on bottom of vessels. No dead space and easy to clean.


BrewstandV3.JPG



Original Post:

First, I'd like to thank everyone here on hbt. Nearly all of my info has come from you guys and I don't think I would have attempted this build without reading about everyone else's.

My original plan was to go with an all electric HERMS system, which used a RIMS style heating element in a pipe and a plate style counterflow heat exchanger(hex). I decided to try RIMS first, mostly because it's more simple and I wouldn't have to change a lot of tubing around during the brew session. I plan on testing the HERMS config in the future. The benefit of the larger surface area for heating the mash is too appealing to pass up.

I'll be using a BCS-460 controller. It's basically a microcontroller box which is accessed via a web-based interface. The BCS will PID control the heater elements via solid state relays. More info about the controller can be found at embeddedcontrolconcepts.com. The BCS will monitor three temperature probes and control two heater elements (rims & kettle), plus the two pumps.

Here's my plan for the mash, sparge, and cooling setup. As you can see in the diagrams, I do not plan on using an HLT. The 5500w heater will serve as an inline tankless water heater during the fly sparge. The benefits of only having two vessels is very appealing to me. I do have provisions to add a HLT later if it doesn't work out.

Parts List (Slightly outdated):

PartsList.jpg
 
Most of the parts are in. Priority #1 is testing the RIMS heater in tankless water heater mode. Because of this, I'm focusing on electrical and control first with the kettles and and sculpture last. The electrical enclosure is nearly complete and I installed a 50 amp GFCI breaker and outlet this morning. I was a little intimidated at first but I was extremely meticulous and everything worked out well.


Here's a pic of the breaker box with the cover off. The 240V 50A GFCI breaker is the big one in the middle right. The 3 wire+ground cable was run down to an external NEMA 14-50R outlet. The 6gauge wire was a pain to work with even though it's stranded.

LoadCenter.jpg



Here you can see the external outlet and the control box. A NEMA14-50 range cord works great for this application since it's mass produced (cheap) and already has terminal rings attached. I tested the system by controlling the little AC fan sitting on top of the enclosure.

TestSetup.jpg



Here's the guts of the enclosure box. It's a fiberglass reinforced plastic enclosure, NEMA 4X (splash proof) rated from McMaster. The main lines from the range cord are 6 gauge going to the large terminal block which is used for the 240V heater elements. The smaller terminal block is for the lower current 120V components (pumps, controller). Once I have the sculpture built I'll run the output lines. I plan on mounting the fan inside the box to cool the SSR heat sinks.

Enclosure.jpg



Here's a pic of the RIMS heater parts. The smaller element is a high density 5500w, slightly bent so it won't hit the sides of the 1.5" ID pipe. The bigger element is a 5500w ultra low watt density that I'll use for the kettle.

Rims.jpg



I plan on assembling the RIMS heater and testing the tankless water heater functionality in the next several days.
 
I am working on mine as well. I am going to feed the stainless pipe heater via gravity and have the March pump on the exit of said pipe heater.

One of us has it wrong I think, not sure who? Does it matter?

You are a lot more organized than me. I have no clue how much I have sunk into this already.
 
Good looking boxer! I've got two. They love it when I brew, since they know that freshly baked treats are only a day away, made up with the spent grain...

Anyway, I'm looking forward to seeing the progress of your rig. The parts list is impressive!
 
I'm curious about the heating capacity of your "tankless" water heater at full volume.

Due to the small-ish surface area of the heating element and short contact time, I can't imagine it will support an 100F differential (ground water: 70F, sparge water: 170F). But, you might be able to achieve that differential if you severely reduce the flow rate through the pipe...

Also, installing a 5,500w element in a RIMS heat exchanger is not normally recommended because of the potential for scorching and excessive caramelization during the mash. Personally, I went with a 1,500w ultra low density element to help me sleep better at night and, potentially, during the mash. :D

Otherwise, looks great! :)
 
Could you run a 5500W heat element at 120V for the mash and switch it to 240V for the Sparge?

I agree with the "tankless" issue. There won't be enough heat to bring it up to sparge temps.
 
I hope all goes well. I would like such a setup but it's not in the cards right now for me. Go for it!
 
The smaller element is a high density 5500w, slightly bent so it won't hit the sides of the 1.5" ID pipe. The bigger element is a 5500w ultra low watt density that I'll use for the kettle.

I agree Lamarguy and Slimer. A 5500 watt high density element in the RIMS heater tube to heat the Strike and Sparge is not practical. He will never hit Strike and Sparge Temperatures and when he recirculates his mash and with the high density element, he will scortch his mash for sure.

He will definitely need something like a 1500 watt low density element in the RIMS heater and heat his Strike and Sparge via other means.
 
I have two boxers also. The first one terrorized me until I got her a friend. They really are great dogs.

You guys pretty much nailed my top two concerns. Since there's no reservoir, I will definitely be limiting the flow. I'm a batch sparger, so fly sparging is new territory for me. From what I've read, fly sparging occurs at rates around 1 quart per minute or even slower. So assuming for a 10 gallon batch, 20lbs of grain, and 36qt of sparge water, I would need to heat 36 quarts of water over a 36 minute time span.

I did some calculations to see if it was feasible. My calculations show a 5500w element should heat 36 quarts of water 100F higher in 24 minutes, ideally:

Code:
Heat capacity of water = 4.184 joules per ml per deg K

36 qt * (946.353mL/qt) = 34,068mL
delta K = 347.039K - 291.48K = 55.56K  (100F rise)

power required = 34,068mL * 4.184 j/mL K  * 55.56 K = 7.92e6 joules

1 joule = 1 watt second so for a 5500w element:

Heat Time= 7.92e6 watt-seconds / 5500 watts = 1440 seconds = 24 minutes

That's assuming 100% efficiency of course. I'm directly injecting heat into the water and the only losses I can foresee is heat loss from the rims piping and through the lines, both of which should be minimal. I don't see any reasons not to expect very high efficiency. Even 80% efficiency would put it at 30 minutes. I may have to sparge slower in the winter, depending on efficiency and ground temps.

I'll try to test it today and prove you guys wrong. :D All I really need to do is teflon up the piping and run a couple wires and tubing.

I don't think I'll have scorching issues. I intentionally made the pipe the smallest diameter possible to allow the fastest flow over the element. I also have the ability to control the individual PID parameters, so in theory be able to throttle the heater.
 
I have two Heat Exchangers and two coolers for my RIMS. Sparge water is heated separately.

Edit - I love my boxer too!
 
I am working on mine as well. I am going to feed the stainless pipe heater via gravity and have the March pump on the exit of said pipe heater.

One of us has it wrong I think, not sure who? Does it matter?

You are a lot more organized than me. I have no clue how much I have sunk into this already.

My thinking was that if you get a stuck mash, the pump could potentinally drain the heater and burn out the element.
 
I don't think I'll have scorching issues. I intentionally made the pipe the smallest diameter possible to allow the fastest flow over the element.

I wouldn't use a high density element in my RIMS heater but I hope it works out for you. You can only flow so fast when re-circulating the mash depending on how fast the liquid filters through the grain bed on top of your false bottom. With the narrow tube you will have a smaller amount of mash in contact with an element that has 150+ Watts/Sq Inch. If you used a low density element it would be half of that Power/Sq Inch and if you used an ultra low density element it would be a quarter of the Power/Sq Inch.
 
The scorching issue will depend on how you control your element. They don't heat up instantaneously and if your control system (controller/ssr) can handle it, a small pulse width will probably keep things in line.

If you're really worried I think you could pulse your setup with water and determine your differential temp change. Then you can set your pulse width accordingly.

I didn't look that close to see what kind of heat sink you have on your SSR, but you'd probably be fine if you add a fan. Just be aware that the smaller your PW the more dynamic power you'll be burning.

Tom

EDIT: Just saw your sparge plans. You'll probably need to have a separate control routine to keep your sparge water at temp :)
 
If this works, it will be the ideal electric based setup for sure. No HLT and perfect temp control of the mash. I agree that the control will have to be different for mash maintain and sparge. During sparge, you'll want to use 100% output to reach the desired sparge temp upon output of the hex. Yes, you'll have to run it slow as hell, but that's fly sparging anyway.

During mash maintain, wouldn't running the element at 50% cycle be the same as a low density? Are the warnings about scorching based on experience or guesses?
 
During mash maintain, wouldn't running the element at 50% cycle be the same as a low density?

Theoretically, you would have to operate the heating element at 25% power to achieve an ultra-low density. But, you may be able to get away with something between 25% - 50% power.

The real issue are the different PID settings (gain, etc.) necessary to achieve both mash and sparge water temperature differentials. Operating the mash with the sparge PID settings will absolutely result in excessive wort caramelization.

Tankless water heaters have a large surface area to water ratio to achieve high temperature differentials. You're better off buying a commercial version than attempting to build one yourself. Don't forget, you'll need to heat the strike water too...
 
Well, tankless heaters also have to accommodate 10x higher flow rate than a sparge would so the power requirement is spiked.

There is theoretically no difference in strike heating times between this method and having an element directly installed into an HLT. In the instant system, the water is flowed in slowly at temp. In a dedicated HLT, cold water blasts in and then you wait for the heat. Either way, you wait.

Even if you built two RIMS style HEX tubes, each designed for their purpose, it's deleting one vessel. It would also allow for a fly sparge on a single tier with only one pump.

Compare the price difference between building a RIMS tube and building an E-HLT. Maybe break even?

All I'm saying is that this isn't one of those ideas that should be discouraged right away. If it fails, all the parts can be reused in the more traditional HLT.
 
Yea I think it's an awesome idea!

You'll really just need a somewhat robust pid and good temp control on your ssr. Once upon a time I set something akin to this up for the fab I worked at while in college. I remember using a pid that had 4 setups that you could cycle through with the push of a button. I wish I remembered what brand it was..... it was blue lol....
 
The BCS allows for finite control over the SSR, but IIRC the duty cycle is at 1 second intervals. If you want 25% power, it cycles 1 sec on 3 sec off. In my head, that would create some major pulsing.
 
If this works, it will be the ideal electric based setup for sure. No HLT and perfect temp control of the mash. I agree that the control will have to be different for mash maintain and sparge. During sparge, you'll want to use 100% output to reach the desired sparge temp upon output of the hex. Yes, you'll have to run it slow as hell, but that's fly sparging anyway.

During mash maintain, wouldn't running the element at 50% cycle be the same as a low density? Are the warnings about scorching based on experience or guesses?

Actually I give the OP an A+ for the concept. I have never scortched my mash because my RIMS uses a 1500 watt low density element. Call it gut feel, but I just can't invision how the mash wouldn't scortch going by a 5500watt high density element. I don't think the flow rate will be as great as he thinks. Don't misread me. I hope I am wrong. Just because Just because I choose to do it differently, doesn't mean it won't work. I have no proof it won't work. It will certainly be an interesting experiment. If he is successful it will be good news for those who are looking to build a simpler system.
 
The BCS allows for finite control over the SSR, but IIRC the duty cycle is at 1 second intervals. If you want 25% power, it cycles 1 sec on 3 sec off. In my head, that would create some major pulsing.

Most run of the mill PID's can do a 2 second period. I'm interested to see if ~200ms will be long enough to cause enough temp variation to scorch. I looked around for response curves for hot water heater elements but found nothing lol ;)
 
Using duty cycle control over the element isn't really want you want, it'll force the element to be active, say, 40% of the time regardless of the temperature. (which is why some PIDs label this function as Manual).

Setting the pulse width (or what we call Output Period in the BCS) isn't really applicable either. That'll set the minimum On time, not the maximum.

So what you're looking for is a way to limit the maximum On time of the output, thereby lowering the effective wattage of element? Cool idea. That feature doesn't exist in the BCS currently, but that's never stopped us before. ;)
 
Keep in mind that I got C's in all my industrial control classes because I didn't care much back then. The tradeoff to short pulse on the SSR is usable life right? I suspect that it would be more practical to run the element at 120v. This should be doable with a DPDT switch where the Hot/Hot is replaced with Hot/Neutral prior to the SSR. I should probably draw it on paper before I write it but I know there are enough experts to correct my mistakes. Sorry/Lazy.
 
Why not use one of the outputs to fire a contactor to switch one leg of the RIMS heater between neutral for low power, and 120 for 240 across element for high power, then the control parameters for cycle time should be in range.
 
Sorry, all that talk about pulse width control and I didn't read that you already had a control unit.....

A good industrial SSR can switch a gigizilllionion times :) It's MTBF is not so much a function of the amount of switches, but more so a combo of heat, steady state and dynamic current, etc... They are cheap, so I probably wouldn't let that hold me back if I were the OP. (just a back seat forum poster on this one)

It is probably easier to use the benefits of 120/240 V operation... Just be careful about throwing switches at full current load...
 
Using duty cycle control over the element isn't really want you want, it'll force the element to be active, say, 40% of the time regardless of the temperature. (which is why some PIDs label this function as Manual).

Setting the pulse width (or what we call Output Period in the BCS) isn't really applicable either. That'll set the minimum On time, not the maximum.

So what you're looking for is a way to limit the maximum On time of the output, thereby lowering the effective wattage of element? Cool idea. That feature doesn't exist in the BCS currently, but that's never stopped us before. ;)

I was hoping I could turn the PID parameters down so that it had a very slow dampened response. Wouldn't that effectively limit the wattage assuming the output period is fast? It's been over 10 years since I took a control systems class and my only real world experience with PIDs is from tuning servo motors.

Triggering a contactor to swap one of the hot legs to neutral is the best backup plan, but i'd rather solve the problem with software. Please let me know when that feature is available :D
 
Swapping the voltage will reduce the response slope so tuning at low output demands would be a lot easier than trying to get a small percentage output at high input power level. Only other trick I can think of is 2 - 2K elements, one in each end of the RIMS chamber, controlled from 2 outputs. When one is fired for Rims use and both are fired for water heating. Build the chamber with a union fitting in middle so elements can be installed and positioned to pass by each other when assembled. Fire each element from seperate leg of the 220 so current flow through neutral is only when one element is fired, when both are in use the current flow is through the hot legs not the neutral.
 
I tested the RIMS heater today and it will definitely meet my requirements as an inline water heater for sparging. I ran a hose to the rims and the output to a plastic measuring cup. I opened the valve to produce a flow rate of 1 quart of water in 50 seconds (1.2quarts per minute, which is more than recommended). The house water was 80F.


RIMS_Test0.jpg


RIMS_Test1.jpg



After I fixed some leaks, I turned the heater full on:

RIMS_Log0.jpg


The rims output temp quickly shot up to 190 after 1 minute with little sign of deceleration so I shut it down, worrying about hitting the boiling point. The slightly flatter graph is the quart measuring cup I used to simulate a HLT. I had about 10feet of silicone hose going to the cup hence the lag. Not quite sure why that temp is higher than the RIMS temp. I did not calibrate the temp probes but they did read the same temp without heat. The three groups of oscillations was me testing the PID functionality set at 120F and 140F. The first test was with the default settings. The 2nd and 3rd was with a shorter cycle period. Wife was bugging me to cook dinner so I didn't get a chance to tune the PID. If I remember correctly, I believe oscillations are due to too much proportional gain, or was that overshoot? I'll need to read up on PID tuning. Once I get it tuned, I'll test the maximum flow rate. That would give me the efficiency of the system and allow me to calculate how it'll do in the winter. Overall, I'm very happy :rockin:
 
Nice man!
Tuning a pid can be a little bit of a pain.... Overshoot can be tuned with the derivative side of the controller.
 
Wow, that is a wild system. The out-of-the box PID coefficients were chosen for systems with a high thermal mass, like most brew systems. Your system has ZERO thermal mass, and the big element makes it VERY responsive to heat. Keep the sample rate (output period) at a minimum.

I'd recommend using the Ziegler–Nichols method, set I and D to zero, and increase P until you get oscillation. Which shouldn't be a problem. :D Make sure to give it enough time to truly oscillate, not slowly converging.
PID controller - Wikipedia

Which way is the water flowing in your picture? If it was left to right, I could see that the measuring cup probe would be higher than the RIMS probe. If not, make sure to check that the RIMS probe is the one that's associated with the RIMS element output.
 
Thanks for the PID pointers. I'll definitely give that a try. The water is flowing right to left and the amplitude of the temp swings (higher swings on RIMS) makes believe that I had the probes wired correctly. But I'll definitely double check. I had some long (10') extensions on the 'psudo-hlt' probe. Would that matter? It was 14 gauge wire so the impedance should have been negligible.
 
I'm keeping up with this thread with anticipation to see how it works out for you. I live in Decatur, work in Kennesaw and brew at my brother's house just over the Paulding County line. I'm in the initial planning phases of going electric for the mash and only using gas for the boil.

Cheers and good luck!
 
Thanks for the PID pointers. I'll definitely give that a try. The water is flowing right to left and the amplitude of the temp swings (higher swings on RIMS) makes believe that I had the probes wired correctly. But I'll definitely double check. I had some long (10') extensions on the 'psudo-hlt' probe. Would that matter? It was 14 gauge wire so the impedance should have been negligible.

Wire extensions won't matter with these sensors, even 100'. I agree, larger swings make sense for the RIMS, and its easy to see that trace0 (gray), which I assume is inside the RIMS, is what is associated with the PID control because you can see that the output switches off when you pass the setpoint.

Actually, thinking about it a little more, it makes sense that the RIMS temp is lower than the measuring cup temperature. Its because there is incoming 80deg water cooling it down.
 
I'd recommend using the Ziegler–Nichols method, set I and D to zero, and increase P until you get oscillation. Which shouldn't be a problem. :D Make sure to give it enough time to truly oscillate, not slowly converging.
PID controller - Wikipedia

So i tried to tune it today and I failed at step #1. I get oscillation at the lowest allowable P gain of 0.01. I also had to make the set point something insane like 2000 for it to even turn on. Any way you could implement that limiter feature? :D

P0.jpg



I could go to a slightly lower wattage element like 3-4kw but that's limiting my strike water heating ability. I did manage to get a somewhat decent response using P=0.3, I=0.15, D=15. The initial ramp from 80F did boil the water some, so I really need to get a lower P, especially if I don't want to scorch using 240V.


PID.jpg
 
One is the RIMS Heat Exchanger temp and the other is a measuring cup with water that brought up to sparge temps from 80 degrees.
 
okay, you need to measure the output temps in a larger vessel. This will average out your fluctuations on the end. If you are using it for instant on demand water for a sparge type applications, put a settling chamber/vessel before the final output. I would probably start with a half gallon and go from there. Make sure there is a fair bit of turbulence and measure the temp on the output side of the settling chamber.

You definitely have more powered than required :)

You can see the affect of the averaging in your first test, the amplitude in temp change is proportionally smaller than measured at the HEX.

The output of the BCS, you could set a PID output to rail between 0 and 100%. Then PWM the result out. Not sure if the processor is fast enough to do it with code, but there are I2C D2A PWM drivers that are pretty cheap. Hec, you can get a PIC with a few on board for a couple of bucks a shot.
 
You might want to start with P = 20, I =1, D =2, the setup parameters you seem to be using are mostly resonding to the "D" factor instead of the "P" factor. FWIW with 8 differnent PID loops that control water, wort, boiler fuel flow, and cooling there have been more than a few hours spent working out the tuning and time variables in the pid loops and a fair amount of practical knowledge gained. With the I & D variables set to one you should be able to slowly increase the P value until it hits setpoint within 60 -90 seconds, then start increasing the I value until you can hit setpoint without serious over/under shoot, then increase D to cut down oscilation. The time interval should be the time between peaks in the output.
 
I am the first one to admit I was wrong. Good job Sizz. I am very curious how the RIMS works with the mash and the high density RIMS element. If it works I will go from my 1500 watt low density element to a higher power high density element to make stepping faster.
 
Just out of curiosity, how close does the hex probe come to the element itself? I'm wondering if it's picking up hot spots just as they leave the surface of the element before it has a chance to integrate with colder water that rode through further away.

Maybe try a longer center nipple to give the water more mix time.

Note: I'm talking out of my ass.
 
If it works I will go from my 1500 watt low density element to a higher power high density element to make stepping faster.

I don't believe the caramelization question has been addressed yet, nor has the question for how to maintain both sparge and mash PID configurations. Sounds like he's currently focused on the sparge temperature/flow question, which, I agree is good news. :)

Have you considered how long it's going to take to get your full volume of strike water up to temperature?
 
Back
Top