Controlling RIMS with a grant

I think I understand how a PID controller works, but so far I have only seen examples with one input (the thermistor voltage) and one output (the heater control). That shoud work fine when the flow rate of the wort is constant, but how do you manage it with a grant?

The way I think a grant works is that the mash tun drains into a grant, and a pump draws from the grant pushing through the heating element up to the sparge arm. But if the pumping rate is not the same as the mash tun drain rate, the level in the grant will either underflow or overflow. I guess most pumps will be able to push wort faster than it can drain from the MT, so there would be no problem with grant overflow. But that leaves grant underflow.

Is grant underflow a problem? It seems like the pump would lose its prime if it underflowed. You could prevent underflow with a wort level detector in the grant. When the grant goes below a certain fullness, turn off the pump. But now you have created a problem for the heating element: without wort flowing by, the heater is not cooled at the same rate, and you could end up scorching the wort. And the poor PID will be very confused.

I guess if you could control the pump speed with a PID controller of its own, you could basically keep the grant level pegged and the pump primed. This should result in a constant flow of wort over the heating element, and the heater PID controller would be happy. Is this how people control a RIMS with a grant?

It strikes me as a sub-optimal system because the heater PID controller, downstream of the pump, would not know what the flow rate is, even though the pump PID controller has that information. Is there a single controller which can monitor grant level and thermister voltage and control pump speed and heater voltage simultaneously?

I don't think most people use a grant.

I was reading the recent BYO last night and they were talking about mash stirring and mentioned RIMS systems and the use of a grant, and the first thing I thought of was, "I have never seen a RIMS on HBT that utilized a grant." In fact, I had never even heard of a grant until I read that article last night.

At any rate, you can put a ballvalve on the output of the pump to limit it's flow. Seems like you would have to tweak it or force pump flow to be lower than MLT flow and then use an auto-sparge valve type set-up to have the MLT flow match your intentionally hampered pump flow, but you could do it.

Any any rate.... like I said above, I've never even see a system with a grant.

Here's the auto-sparge thing I was talking about. http://www.rebelbrewer.com/shoppingcart/products/Auto-Sparge.html

If you put that in the grant and had it controlling the valve that was allowing drainage from MLT to grant, then you could just throttle back the pump with the ball valve on it's output and let the auto-sparge match the rate into the grant for you.

Not sure if this applies but often when manipulating pump flow, a bypass line back to the source is an option. Basically have a tee on the outlet side, one line goes where it normally did and the second line goes back to the grant. Then you could stop flow into the tun and just send it all back to the grant if you wanted, or vice versa or anywhere in between (valves required of course). Just thinking out loud, sorry if not applicable.

SpanishCastleAle said:

often when manipulating pump flow, a bypass line back to the source is an option. Basically have a tee on the outlet side, one line goes where it normally did and the second line goes back to the grant. Then you could stop flow into the tun and just send it all back to the grant if you wanted, or vice versa or anywhere in between (valves required of course).

Click to expand...

This is brilliant. It means that the flow past the heating element is nearly constant regardless of the mash tun drain rate. There might be some effect of different shunt ratios on the load of the pump causing the pump output to change but it would probably be minor.

The shunt valve could be controlled by a PID controller based on the grant level, but there would be absolutely no need to "feed forward" the flow rate to the heater control PID.

The only drawback I can imagine is that the pump might use more energy than it necessarily would have to if the pumping rate were matched to the MT drain rate. But that seems a small price to pay for stable heater performance.

I like it!

You might have some problems trying to implement this, but I'm not sure.

I have never seen a PID controller that uses anything other that temperature as in input. They might exist, or you might be able to turn the input from some other kind of sensor into something that a PID temp controller can use, but there is still the issue of how multiple PIDs might interfere with each other.

"PID" is an acronym for the name of a control algorithm that uses parameters for "P"roportional, "I"ntegral, and "D"ifferential information. These parameters have to be set specifically for the system the PID algorithm is controlling.

A simple thermostat is a REactive device. When the temp falls below a set-point, the heat kicks on. When the temp goes above the set-point, the heat turns off.

A PID temp controller is a PROactive device. The values of the P,I, and D parameters allow the PID to anticipate when a temp drop is about to happen and it will boost the heat before the temp drops. It's an intelligent thermostat.

The problem with a bypass from RIMS output back to the grant is that the PID that is controlling the heat will not be monitoring a simple system. When wort is going back into the grant from the output of the RIMS, that wort is going to make multiple passes through the heater and thus will warm up faster than the actual mash in the tun. The temp PID will then think that the thermal mass of the system is smaller than it really is and won't heat as much.

Long winded, but I think two controllers set up like this will confuse the temp controller and you'll be off on your mash temp.

If you used a device that just regulated input to the grant, then the temp PID won't be feeding back onto itself and making the thermal mass look smaller than it really is.

I think you're over-thinking this, unless the desire is for a totally automated system.

My belief is that if you allow the grant to fill to @ 1/2 capacity, then begin to pump, it takes all of about 2 minutes to match flow rates to a reasonable amount, and in 10 or 15 minutes you can take another look to make sure you're not over-feeding or under-feeding the grant.

As long as you can observe the level in the grant, there really isn't that much to it.

Walker said:

I have never seen a PID controller that uses anything other that temperature as in input. They might exist, or you might be able to turn the input from some other kind of sensor into something that a PID temp controller can use

Click to expand...

They absolutely exist. PIDs are used to control all manner of variables. If you can control a system electrically and if you can measure its state electrically, you can control it with a PID. That doesn't necessarily mean that the system will be stable, for many reasons, but not least of all because (as you pointed out) there may be other inputs that drive the system state.

Waker said:

The problem with a bypass from RIMS output back to the grant is that the PID that is controlling the heat will not be monitoring a simple system. When wort is going back into the grant from the output of the RIMS, that wort is going to make multiple passes through the heater and thus will warm up faster than the actual mash in the tun. The temp PID will then think that the thermal mass of the system is smaller than it really is and won't heat as much.

Click to expand...

I think you are correct about this risk. But think of it this way: turn on the pump controller first, without turning on the heater. The wort will circulate at the mash tun temperature (minus heat sink effect of the RIMS components), and the grant level will seek the level set by the controller. Once that level is stable, you will have a constant rate of flow through the heating element and a constant rate of flow from the heating element into the mash tun.

At this point, turn on the heating element controller. It will start heating wort, and the wort will recirculate both through the mash tun and the grant. The contribution of heat from the shunt will be larger than the contribution of heat from the mash tun, so the heater output temperature will rise faster than it would have had there only been the contribution from the mash tun. The mash tun will fill with heated wort at exactly the rate that it drains, which is much slower than the rate that the shunt fills the grant with heated wort. What you'll see is that the temperature coming out of the grant will be 90% from the shunt and 10% from the mash tun, and therefore should be even more stable than had it only been from the mash tun.

I think this system will be stable. I am not a process engineer and I don't even play one on TV, but I think this is worth a try.

Are you recommending regulating mash tun output as an alternative to using a grant? That might be less complex than regulating the grant level. Something to think about.

I had never even heard of a grant until I read that article last night.

Click to expand...

I am a grant user.


Cheers,
ClaudiusB

yes, I was suggesting regulating the MLT flow into the grant, much like a fly sparger limits HLT flow into MLT.

Restrict the pump output to something reasonable, yet lower-flow than the MLT output. You have now created a constant flow past the heating element.

Then use the mechanical auto-sparge valve to keep the level in the grant constant by starting and stopping the output of the MLT. One less electrical device at least. It's just flow rate matching that was mentioned earlier, but automated mechanically.

You might even be able to rig up the autosparge thing to do what you were talking about.

Probably cheaper than buying a special PID, level sensor, and solenoid valve.

There is a nice advantage to the bypass idea.

If you ever get a stuck sparge or a clog in the MLT output, the bypass would end up fully opening and flowing back into the grant, which would prevent the pump and RIMS from running dry.

It removes the potential to dry-fire the element.

ClaudiusB said:

I am a grant user.

Click to expand...

Good Lord! I just looked at your brewery pics.

I just made a grant for my RIMS system also. My system is so compact that my pump would put too much suction on the grain bed. Thereby compacting it and making it impossible to recirculate. Don't have a picture of it installed, as I just finished making the grant. But here it is.

It's just a little 2qt Bain Marie pot with quick disconnects.

In theory, it should work fine great and I will be able to control the flow via ball valves on my pump/mlt.