@crutschow: I like your reasoned responses in this thread. Clearly,
@enli does need some sort of battery-powered heater to maintain a constant "calibration" water temperature of either 140°F ±4°F or 170°F ±4°F, although not both temperatures at the same time. You could use
two water baths, with one maintained at 140°F and the other at 170°F, to save time. I would seriously consider this alternative. However, there
must be some way to avoid an excessive loss of water temperature, without expending an excessive amount of battery power, while moving the calibration rig from kiln to kiln. The OP,
@enli hasn't said how much "time in the field" is expended to move the calibration rig from kiln-to-kiln and to record the results, although the following statement provides a "clue":
Need to maintain test temperatures for at least 30 minutes (for testing 3 or 4 probes at each site).
This may be inclusive of the time needed to move from the calibration rig, located at a particular kiln, to the "control room" where presumably the permanently-installed kiln thermocouples are read and their value recorded. Then whoever is certifying the calibration moves the calibration rig to the next kiln, makes sure the water baths are at the correct temperature and then returns to the "control room" to record and document the calibration results.
This sounds like a two-person job for a team equipped with walkie-talkies. One person stays in the control room to record readings, the other stays in the field to pull and insert kiln thermocouples into the calibration rig. Sure, one person could do both jobs, running back and forth as necessary to beat the thirty-minute "cool down" period of the calibration rig, but it would be better to have the calibration rig
not cool down at all between individual kiln calibrations. Hence the need for a heater and some good thermal insulation.
The two main ways to minimize heat loss, and hence temperature degradation, are either to use a massive amount of thermal mass (larger water bath), or better thermal insulation, or both.
All of this is measurable, and to some extent controllable, especially on a small scale. The science of it is called thermodynamics. And as
@(*steve*) says in his tag line: Thermodynamics... it's not just a good idea, it's the Law!
Having had some experience with temperature control, I must say that I would
not suggest increasing the mass (volume) of water for this type of calibration measurement. Water has a large thermal heat coefficient (4.184 J/g °C), meaning it takes a relatively large amount of energy to change its temperature, so it is better to heat a smaller volume of water (quickly) and to increase the thermal insulation to minimize heat loss. Even so, a good PID (Proportional, Integral, Derivative) control algorithm is necessary to heat the water quickly to the desired temperature with zero overshoot, i.e. critical dampening of the temperature control algorithm.
Why critical dampening? Because your (hopefully) excellent thermal insulation will slow down any recovery from overshooting the calibration temperature. This is a typical problem in trying to control the heated temperature of a well-insulated environment: the environment heats up a lot quicker than it cools down. If you have lots of power available for heating the water, you could get a faster response by increasing the heat losses, for example with less thermal insulation or by means of a thermoelectric (Peltier device) cooler that can either extract or provide heat. I wouldn't go there with a portable battery-powered water bath, unless you don't mind pulling around a large cart loaded with batteries. Peltier coolers are not very efficient.
Thinking the water bath would be a 12 to 20oz container, but this is just one of the variables that can be changed for maintaining accurate temperatures.
My "gut" feeling (after looking at a 20 oz Gatorade bottle) is 12 oz looks about right but 20 oz is waaay oversized for
@enli's application. Your really need a vacuum-insulated container, such as
this over-priced 12 oz Starbucks vacuum insulated coffee tumbler: Buy two of them so you can have both temperature references handy.
From here on it gets a lot more complicated, depending on how accurate your results must be. You stated that the "gold standard" for your calibration test was a digital scientific thermometer, which presumably would be inserted in the water bath to verify its temperature, and perhaps to actually control the water bath temperature by means of a negative feedback control of an electrical resistance heater immersed in the water bath. Sounds like it's getting a little crowded inside the water bath chamber!
Each of the thermocouples you remove from the kiln has a thermal mass and a specific heat capacity that will literally begin to suck heat out of your water bath as soon as the probe is inserted. it is up to your immersion heater to resolve this problem by adding back the amount of heat that was initially lost because of probe insertion. Unfortunately this is NOT a simple problem with a simple solution. Still water is an effective insulator. Just because the heater is on does not mean the entire volume of water will be at the same temperature.
There are two ways to accomplish a homogeneous water temperature profile: time and stirring. With time, any quantity of heat, added at a point to a volume of anything, will eventually disperse evenly throughout the volume, provided there are no thermal losses or gains at the boundaries. This process can take hours, or even weeks, with good thermal insulation. To speed up the temperature homogenization process, the next step is to stir the liquid. This is very effective. Even more effective is to circulate the liquid past or through an electric immersion heater.
So there you have it: the outline of a design for your kiln thermocouple calibration rig. If you decide to go this route, I recommend using a peristaltic external aquarium-type pump to circulate water bath water around an immersion heater. This does not have to be a commercial immersion heater like the kind sold for preparing a cup of tea or instant coffee. Any resistor of about ten watts dissipation capacity with suitable resistance will work. Use silcone sealer to coat and insulate the under-water soldered wire connections.
I would also consider designing my own microprocesser-based heater control to minimize the "warm up' time of the water bath. Try to obtain a stead-state electrical consumption of around ten to twenty watts during warm-up and perhaps a tenth that much after equilibrium temperature has been reached. These are just rectal extractions from my experienced arse, so it would pay to see if they can be made practical for you. I would start by inserting thermocouples inside my vacuum-insulated container, adding hot water, and recording the temperature drop as a function of time, say every minute for at least thirty minutes. Record either the weight or volume of water used. After sealing the container, shake it once every five minutes to evenly distribute the heat losses. This data will give you some idea of how much energy you need to add to the water to maintain the set-point temperature for thirty minutes.
Always remember to cap and insulate the mouth of your vaccuum-insulated container to avoid excessive heat loss by this mechanism.
And if you want some really good professional information and advice on all things related to temperature measurement, control, and calibration, my advice is to seek help from
Omega Engineering Inc. I have used this company for more than fifty years and trust their advice and products. As mentioned before, it helps to know the rudiments of thermodynamic engineering before attempting to engage them in conversation, but their online catalog is a big help in that area too.