Anvan wrote:So thanks Ray, and I look forward to seeing what's next.

I've been talking with DrGary about installing a TC or RTD sensor into the bottom of the portafilter basket. As you correctly point out, the temperature differential between the boiler itself and the group head affects the water quite a bit.

It's like heating water in a pot to a specific temperature (not much of a problem), and then dumping the water onto your kitchen floor and hoping to be able to always achieve the exact same temperature when the water is on the floor.

Nevertheless, I think it would be interesting to install a TC into the basket and start taking readings. Since the La Pavoni I use does have the PID controller for the boiler, I think it would help highlight what sort of temperature drop there is. My thinking was to pull 10 shots in a row (which would [should] steadily increase the group head temperature). A graph of this might help show whether or not the group head reaches a plateau with regard to its temperature and whether or not this might have a stabilizing effect when it comes to how much of a difference there is between the boiler and the group head (i.e. water temp when the water is actually doing the extraction).

My gut feeling, after using the machine for a number of months, is that once the machine is up to temperature and the group head is hot (via pulling a few shots), there will probably tend to be some amount of stabilization that naturally occurs. This may be more true due to the PID controller, since it keeps the boiler temperature pinned down by about .1 degree. This being the case, I think the boiler (and the water) may tend to help regulate the group head temperature (once it's up to temp). If that's the case, then I would think the temperature drop that occurs from the boiler to the group head would be fairly predictable.

Right now I'm rebuilding the suspension on my motorcycle, so the La Pavoni "group head TC" project is down the list a bit

Anyway, I just thought the chart might be helpful to people who have a gauge and may have some interest in where the boiler temperature is. If you want to use a calculator to compute the boiler temperature instead, don't let me stand in your way

Ray

People who have PIDed boilers frequently report that there are differences in temperatures throughout the boiler; i.e. the boiler is not at temperature equilibrium, since the heater cycles. The steam tables give the pressure/temperature relation of steam at one spot or for a system at equilibrium. Your readings do not accord with these table because you have picked a spot to measure the temperature that does not correctly reflect the average.

I was involved in replacing a pstat with a PID controller about three or four times. In each case, I had no trouble finding a spot for the the TC where the temperature reading most closely approximated the steam table prediction for the pressure gauge reading. The reason I did this is simple -- the pressure is highly uniform throughout the boiler and reflects the average temperature, while the temperature gradients in the boiler make actual temperatures at any spot variable.

Bottom line, you can monitor boiler pressure anywhere; but you need to be careful when picking a spot for temperature. If you get too close to the element, you will regulate the temperature for the water mass around it and leave the rest of the boiler cold. If you are too close to an uninsulated wall, you will be regulating the air outside the boiler and overheat it, etc etc. If your temperature readings follow the pressure gauge readings, you know you have the sensor at a good spot for accurate control.

It's much harder to do a good job with a brew boiler PID than a steam boiler one, precisely because pressure is such a perfect indicator of what is happening overall.

Thanks Ray, that table is useful and informative. As a Pavoni user I appreciate all your posts. However, I agree with Jim, that one can use a steam-pressure table to get the same info.

Be aware that you are reading gauge pressure, so you need to add about 1 bar to compare with the absolute pressure that you see in the steam tables. (1.01 bar at sea level, .98 bar at 800 ft altitude, depends slightly on the weather as well.) When I did that it looked to me like your results are pretty consistently 3 or 4 degrees below the expected values, which could be due to TC placement and perhaps the dynamics of your system. (I assume you are bleeding off a little steam while reaching a steady temp/pressure state.)

Note - at around 5000 ft altitude, the atmospheric pressure is only about 0.84 bar, which would work out to about another 4 degree drop in temperature at a given gauge pressure. So if you live in Denver or Albuquerque, Ray's table is pretty close to what is predicted using a steam table.

- Pat

Edit - Original post said 1.03 bar at sea level, 1.0 bar at 800ft, which is off a bit and I corrected that.

another_jim wrote:People who have PIDed boilers frequently report that there are differences in temperatures throughout the boiler; i.e. the boiler is not at temperature equilibrium, since the heater cycles. The steam tables give the pressure/temperature relation of steam at one spot or for a system at equilibrium. Your readings do not accord with these table because you have picked a spot to measure the temperature that does not correctly reflect the average.

I was involved in replacing a pstat with a PID controller about three or four times. In each case, I had no trouble finding a spot for the the TC where the temperature reading most closely approximated the steam table prediction for the pressure gauge reading. The reason I did this is simple -- the pressure is highly uniform throughout the boiler and reflects the average temperature, while the temperature gradients in the boiler make actual temperatures at any spot variable.

Bottom line, you can monitor boiler pressure anywhere; but you need to be careful when picking a spot for temperature. If you get too close to the element, you will regulate the temperature for the water mass around it and leave the rest of the boiler cold. If you are too close to an uninsulated wall, you will be regulating the air outside the boiler and overheat it, etc etc. If your temperature readings follow the pressure gauge readings, you know you have the sensor at a good spot for accurate control.

It's much harder to do a good job with a brew boiler PID than a steam boiler one, precisely because pressure is such a perfect indicator of what is happening overall.

My RTD sensor is in close prolixity to the heating element, that's for sure.

When you talk about a steam boiler PID, are you talking about mounting a PID sensor in the steam chamber (e.g. the steam part of an HX machine or something)?

Ray

Your pressure gauge is giving an accurate reading of the average boiler pressure. If you can push your RTD further into the boiler or pull it further; you'll usually be able to find a spot that tracks the pressure gauge reading closely.

Your numbers aren't far off of steam tables at sea level.

For example, the "Steam Tables" app on my Andriod phone shows
1.40 bar abs, 0.40 bar gauge for 228 deg F (compare 0.48)
1.94 bar abs, 0.94 bar gauge for 246 deg F (compare 1.07)
~0.1 bar off in both cases, in the same direction

I don't know your altitude or barometer when you made the measurement, as others have mentioned. There is also the reality that most of the gauges on espresso makers are horridly inaccurate. Accuracies of commercial gauges are typically 2-3% of full scale, probably twice or worse than that for the "cheapies" used by home equipment manufacturers, so 0.1 bar (~3% of a 0-3 bar gauge) off isn't unexpected.

I don't know how you calibrated your PID. Being off by a couple degrees there isn't out of the ordinary. Working backwards with the same app
0.48 bar gauge for 231.7 deg F (compare 228)
1.07 bar gauge for 250.3 deg F (compare 246)
~4 deg F off in both cases, in the same direction


To echo Jim's response, the steam tables apply to the temperature of the steam/water interface. If you're far enough away from that interface that you're measurement is significantly impacted by heating (or cooling, through the walls, for example), moving the probe might help get your measurements more consistent with the tables.

Then again, I only use temperature for a relative thing -- "Two degrees cooler than normal, for me" and don't much care about the about the absolute. Since I don't have an exact copy of the roaster's equipment, them saying "201.2 deg in a blah-blah basket dosed at 19 g for 27 seconds and 1.25-1.5 oz yield" usually leads me astray from trusting my own ability to dial in a shot from my standard grind/dose/basket/temperature starting point.