Electronic RH sensors

Can electronic relative humidity sensors stand up to the elements in dry kilns? May 2, 2001

I'm interested in finding out more about electronic relative humidity sensors used in dry kilns. Exposure to the acidic, high humidity, high temperature environment of a dry kiln is always an issue with these types of sensors.

Forum Responses
I do not believe that anyone makes RH sensors that will stand up to normal kiln temperatures and humidities. Lignomat does make an EMC sensor (looks like a white piece of blotter paper) that is widely used. Brunner also has a similar style sensor.

Gene Wengert, forum technical advisor

From the original questioner:
I think Honeywell might now be making one that works for a reasonable length of time. I think you're right that the cellulose-based ones are questionable with respect to performance and longevity, unless there's been some recent advances.

I agree with Gene. We have never found an RH sensor suitable for a dry kiln. We use them often in pre-dryers and in applications where conditions are milder and fluctuate less. The electronic sensors, and we just looked at Honeywell, are least accurate where you need to be most accurate when drying green lumber. Most are not reliable over 80% or under 20% and, when drying oak for example, that is where you need to be dead on. Finally we have had problems with them drifting out of calibration when the rated temperature is exceeded. We do use them for people who only dry at low temperatures and who only dry air dried and do not have critical final moisture content requirements. If that is you, then go ahead and try it. A wet bulb is my preference.

We use Lignomat temp and EMC sensors in six of our kilns and in one pre-dryer used for white woods. We also have four pre-dryers using an old style Honeywell sensor. The Honeywell sensors are never in calibration and most of the parts are no longer available. The Lignomat sensors are the most accurate that I have ever come across, and the simplest to repair if there ever is a problem. We are also running four kilns with dry/wet bulb systems which are good too, but difficult to calibrate, troubleshoot, and repair.

There are several companies who offer the EMC sensors and they work well. Lignomat, a USA company, makes good stuff. There are lots of companies in Europe doing it. As for calibration of wet bulb and dry bulb controls, the old mechanical ones were tough, but with the newer ones using an RTD, they are really easy to check and calibrate and it is rarely necessary.

Omega engineering makes a handheld temperature/humidity/dew point meter. The model number is RH70 and the temperature range for the humidity sensor is rated for -5F to 175F.

I like the $29 handheld from Radio Shack. Extremely accurate and durable.

The only two problems with DB/WB sensors is getting the water to them all the time and getting enough airflow. I have suggested putting a small fan on the WB sensor. Turn it on and off and watch the reading change depending on whether the fan is on or off. Often the entire kiln runs better with adequate air across the WB!

Gene Wengert, forum technical advisor

From the original questioner:
What I'm hearing is:

a) No one seems to definitively know which ones really work and which ones don't for this particular application.

b) The ones that are being used aren't reliable and/or accurate in some key situations like low RH or high RH conditions.

c) Other varieties being used (cellulose-based) are known to have problems related to long-term exposure to the harsh industrial environment of a kiln.

d) At the end of the day, it's best to buy a cheapy that's hand-held and the solution is to keep your hand stuck inside the kiln for the duration of the schedule and record the numbers with your free hand.

Sounds like a good research project in the making. When one does a Google search on RH sensors, there appears to be a whole swack of them out there. It would be nice to throw a half dozen different ones into a red oak kiln over a year, and see which really work after 12 months. I'm surprised this hasn't been done yet.

We have been using RH transmitters for almost four years. We have installed them in pre-dryers, DH kilns, vac kilns and conventional kilns.

We have used three different brands. The first was Ohmico. It was an import from Europe. It drifts but is easy to calibrate. The second is from Elan Technology. It's stable and, if it becomes contaminated, you rinse it off with distilled water. The third is from Rotronics. I like it because it is very stable and the sensor is well protected inside a teflon filter.

Wet bulb measurement is often very inaccurate. When the air is blowing over the top of a charge and hits the WB in the back, the controller might respond to the dry air by spraying. When the air is going through the charge and then to the WB, the controller responds by venting. That's why you see an oscillation in DB in one fan direction. So we use two transmitters. One on either side of the charge.

Accuracy is typically +/-2% from 10 to 100% RH. Look at your EMC chart in the DKOM and you'll see that a wet bulb can't come near this accuracy.

Temperature operating range is up to 160. After 160, accuracy isn't guaranteed. Above 160, what difference does a little inaccuracy make?

If you find an accurate portable meter for $29, go buy a lottery ticket. It's your lucky day.

Today we installed four in two new kilns for a customer that has around 20 or 30 of them.

We have compared 8 Radio Shack $29 units against an $800 sensor and found that none were off by more than 2% RH. Before you make your statement again, spend $29 and see if this is not true. I am sure that once in a while they have a bad unit, but it seems that most of them work great.

Your comment about WB being inaccurate is wrong. They actually are the basis for humidity measurement and have been since 1909. Regarding the EMC, the chart in the handbooks is for 1 F increments; if you need closer values (but in kiln drying we do not), there is a formula that you can use to get more precise. But, note that RH to EMC conversions are equally inaccurate. The WB/DB is the most accurate basis for EMC and RH measurement that ordinary people have available.

You also seem to imply that the WB temperature varies through a kiln load. This is not true, unless there is heating inside the load. Hence, only one WB sensor is needed in a kiln (unless the kiln is long). The DB, EMC and RH does vary from side to side, but not WB. That is why you have to use two sensors.

The literature for every RH sensor I have looked at for use in dry kilns has a statement that it is not to be used in acidic atmospheres. Apparently you have found two or three that do not have this restriction.

Gene Wengert, forum technical advisor

Gene is right. RH is a derived value (calculated as a ratio of vapor pressures) based on wet bulb measurement. It goes back to the basis of psychrometrics, which was developed by Willis Carrier nearly 100 years ago. The easiest measurement to get and be sure is accurate is wet bulb and dry bulb. Over the years people have developed chemicals or devices that closely mimic the values you get with wet bulb and dry bulb but they cannot be totally accurate over the whole range. Even the EMC sensors, which use wood or a paper product, attempt to mimic the values one would get by measuring wet bulb and dry bulb. Gann patented the use of wood chips so most of the other companies use some sort of paper. The wood chip or paper is held in a clamp and the moisture content is read by a moisture meter just like any moisture meter. This is the EMC and it is obvious there can be some inaccuracies in that. However, people get way too anal about this in dry kilns. You usually have a pretty good margin of safety.

I didn't say wet bulb was inaccurate. I wrote "is often". I find cruddy wicks, RTD's nearly in the bath, the air flow is rarely correct (especially in DH kilns) and there is always the problem with reading WB before the charge in one direction and after the charge in the other.

There is actually a more definitive method then either dry and wet bulb or RH wafers. It is dry bulb and dewpoint method. A study has been performed comparing many bands of RH wafers and dry and wet bulb method to the dry bulb and dewpoint method. If the wet-bulb is maintained properly it was more stable and accurate than RH wafers. Just because some people do not maintain the wet-bulb properly does not invalidate the method. The need for great amounts of airflow over the wick is overstated. Usually, the problem with wet-bulb is exposing it to a surface that is not the same temperature as the kiln air. Such as a cold wall in the winter or in direct line of sight to the heating coils. Again, though, this does not invalidate the method.

This is true. I have connected chart recorders to RH transmitters in kilns alongside WB's with recorders. The WB's are amazingly close at times despite less than laboratory conditions.

Does anybody see their DB fluctuate in one fan direction but not the other?

I am assuming the following is your question? "Wet bulb measurement is often very inaccurate. When the air is blowing over the top of a charge and hits the WB in the back, the controller might respond to the dry air by spraying. When the air is going through the charge and then to the WB, the controller responds by venting. That's why you see an oscillation in DB in one fan direction. So we use two transmitters. One on either side of the charge."

It sounds like there are many problems present. Is there any baffling on top of the stack? There should be.

How dry the air is because it does not go through the stack does not change the wet-bulb reading - it stays the same from the point of entering the stack to leaving the stack. Is there enough water present at the wet bulb wick or is it exposed to the heating coil? Both of which would be interpreted by the controller that the air is dry.

The dry bulb should not be oscillating within one direction of airflow. It should oscillate as the air direction is changed. If it does oscillate within one direction something is wrong and a close inspection of the operation is needed. (In other words, I can not think of a reason from this distance.)

This kiln is hypothetical. Suppose heat and fans are above a kiln charge. It's a package kiln. Let's call the fans blowing toward the door "forward". When the fans are in forward, air goes into the door side of the charge. The wet bulb "sees" air that has lost some heat and has picked up some water. When the fans reverse, heated air goes down the back of the kiln and directly on the WB. Are conditions the same?

The process is adiabatic, so the wet bulb will be the same. If there is a variation it is because of other factors--leaks, etc.

Which process is adiabatic?

It is correct that the process is adiabatic, so the WB will be the same. Adiabatic means without loss or gain of heat energy. (That is why I stated that the WB was the same "unless there is heating inside the load.") When air passes over the heating coils, the process is not adiabatic. Venting is also not adiabatic. But once the air enters the load, the process is adiabatic until the air exits (unless there are center heating coils).

Bad WB readings are the result of bad operation of the WB--not enough water, too hot water, poor air flow, poor wick, etc.

The wet-bulb reading (not the actual WB value, but the indication on an instrument) will vary due to different airflow across the bulb in the two fan directions. Often, when the air is going upward in the plenum, the flow across the wick is poorer than when the air is going downward. Also, there is one kiln design that has the vent fairly close to the wet-bulb, so that the cold vent air affects the WB reading when the fans blow in one direction. In this case, any sensor would be affected.

Dry-bulb temperatures can fluctuate in one direction because of the plenum size, vent locations (and poor mixing of the vent air with the kiln air), heating coils that are not balanced (trap on each one) so that one bank heats more quickly than the other, direct exposure to heating coil radiation (IR radiation), or poor sensor location. Perhaps there are other reasons, too.

A related problem occurs with two DB sensors that are switched on and off (reversed) when the fans reverse. Such reversing of the sensors is done with a relay and often the contacts wear. With resistance sensors, we are dealing with very small resistance differences being large temperature differences. In computer systems, the sensors are not switched, but the ones that the computer uses for control (the active sensors) will be switched. In this case, there is no relay switching; the data are switched in the computer.

In older control systems, whenever there are questionable and strange readings, the first question is whether it is the sensor giving a false reading or if it is actually the air temperature that is changing.

Back to WB readings, the theory (thermodynamics) of the WB is well described mathematically (Byrd, Stewart and Lightfoot). In this text, the affect of velocity is also described. At approximately 600 fpm across the bulb, changes in velocity have little affect. However, if you get under 400 fpm, variations in velocity produce variations in cooling. Another important factor is the temperature of the water--it cannot be too cold or too hot. This is especially critical in units with large water boxes. Also, the WB data is based on a muslin wick. Some of the wicks used today are not as absorptive as they should be. In fact, I encourage everyone to wash their wicks once (no soap) to rinse out the sizing used in many fabrics. Sizing makes your new shirt look like a new shirt to everyone else, but interferes with evaporation!

Gene Wengert, forum technical advisor

In a closed system, I suppose that the wick might be heated more by air temp in reverse but cooled more by WB depression, while in forward it's heated less but cooled less. I still don't see how you can call "the process" adiabatic. Even if it is adiabatic in a closed system, how does that apply to a kiln where you may be spraying or venting plus you're heating wood and you are losing heat from a structure?

The process we are looking at is the space from just before when the air enters the stack to when it just exits the stack. There is no loss of energy. The total energy in the air is constant. The dry-bulb temperature is lower but the energy involved in lowering the temperature is contained in the moisture that the air gained. So the energy of the air and the energy of the moisture in the air adds up to the same energy you start with. That is what is meant by adiabatic.

If the sensor is not positioned where it receives the air coming out of the stack before it is affected by other factors, then the reading is in error.

So, no matter which way the air is flowing, the total energy is the same and the wet-bulb is constant.

That is true if every unit of heat which doesn't exit has been used for vaporization.

The key is ENERGY and not HEAT--or hot air, cooled air, or vaporization heat. Adiabatic processes are concerned about ENERGY, in any form. Perhaps this idea will help your understanding.

Gene Wengert, forum technical advisor

Every unit of heat that enters the stack of lumber leaves the stack of lumber. It doesn't get "used for vaporization". Some of the sensible energy is changed to laten energy and it does it by moving along the wet bulb line towards saturation, but may never reach it. This is the basic psychrometrics.

There is a huge thermal transfer between liquid and vapor states. But that doesn't matter since we have already established that there is no net loss due to vaporization. What about the heat that raises the temperature of water that didn't vaporize and the heat that went into the wood?

An adiabatic process is where there is no energy added or subtracted from an outside source. An outside source could be a boiler or could be venting. But the air going into a stack does not lose energy to an outside source--the energy goes into the wood or into evaporation. (Incidentally, the energy used for heating the wood is quite small, except during initial heating.) A dry kiln has a non-adiabatic process from one side of the fin heating pipe to the other side, or between the vents, or right around the fan motor which is adding heat.

The key is that energy is not used or disposed of in the lumber stack. All the energy going in or coming out is accounted for within the stack of lumber.

Gene Wengert, forum technical advisor

It's easy to understand the theoretical process. Application to the real world is the hard part.

I have in front of me charts from when I considered this question a couple years ago. The charts note 5/4 cherry. One chart shows a six to eight percent shift in RH when the fans change direction for about the first three days. Then the difference when fans change is minimal. The difference recorded could have been the difference in transmitters. This is the point where minimal energy is left behind in the stack. It stays this way until the last four or five days when the RH again starts shifting up and down with the fans changing direction. It was about a 6% shift downward when the DB was 130 and the fans reversed.

When lumber is dry (under 15% MC), the RH difference from side to side is very small. You can see this by assuming a 3F DB drop as the air passes through the load (hardwood lumber) and use a constant DB. So, you might have 160 F and 115 F entering and 157 F and 115 F exiting. There is a small RH and EMC difference between these two values. Of course, if you use the same calculation when at 130 F and 122 F (8 depression), you will see that the exit at 127 F and 122 F has a large RH difference.

If your sensors are removable, switch them with each other. Now see if the humid side has also changed. Also, I have seen one kiln where the RH was a calculated value (the sensor did not adjust for temperature in itself, but relied on the computer to do it) and the programmer forgot to change the DB used in the calculation for the sensor. This gave a false DB in one direction, but not the other. This is also a problem with DB/WB sensors that report or record depression and EMC or RH, rather than just the temperatures.

Gene Wengert, forum technical advisor

You don't have to switch sensors. Just turn off the fans and leave the door closed.

It is very easy to develop temperature gradients in a dryer with the fans off. Temperature gradients (since the WB is constant) will mean RH gradients. Therefore, TURN THE FANS ON.

Gene Wengert, forum technical advisor

The stack of lumber is essentially a big wet bulb wick and just as you need airflow over the wet bulb wick in order to get an accurate reading, you need it over the lumber. If you think of the lumber as a big wick, it might be easier to understand what is going on thermodynamically.

We were talking about comparing sensors. We mount an RTD next to each RH transmitter so temperature gradients are obvious.

If we were talking about a conventional kiln operating in a cold climate, you might think of the door as a big condenser to make it easier to understand what's going on thermodynamically.

Total accuracy is a moot point anyway. We replaced the control system on a NYLE last week. The wet bulb RTD was in a wick stained brown. The RTD was almost in a water box crusted with corrosion and filled with warm water. The operator had learned to deal with it. The new controller uses one RH transmitter.

The comments below were added after this Forum discussion was archived as a Knowledge Base article (add your comment).

Comment from contributor A:
I've dried lumber for 20 years in steam kilns, direct-fired kilns, slope grate green fueled kilns, fluidized bed kilns. I have used one control system that has worked with every one of them except the fluidized bed, which did not have it installed. I use a RH derivitive system developed by drying technologies as the primary shutdown parameter. I also have the new Accudry system installed on one of my kilns. Beleive me, the RH method works more consistently than any other. I have Delta T, wet bulb drop, RH, wet bulb set point and max time as choices. The RH is based on a chart using wet bulb/dry bulb ratios. Like I said, I am not a professor or engineer, but I dry lumber every day of the week.

Comment from contributor B:
We are working in the field of Tea processing where %RH plays a very vital role. Our experience with RH measurement is also the same as yours that polymer sensors fail to function because of the deposition of tea extracts (polyolefins and tannic compounds). We have ultimately switched over to a conventional dry-bulb/wet-bulb method with temperature measurement accuracy of 0.2 degrees Celsius using RTD elements.