Question: Quantitative analysis of wood moisture

MarcioPamplona is asking a question about spectrometry
Follow this topic

by MarcioPamplona | March 19, 2018 09:55 | #15971

I need some advice on how to design a system to measure the amount of moisture in wood.
Measurement by reflection, directly on the surface of the material.
Do you have to calibrate for each type of material or can we directly measure the amount of water molecules in the material?

Thanks in advance


Hi, are you interested in using a spectrometer? (Just guessing from tags) I'm not sure about if that's possible, but happy to try to think through it with you. From how far away?

I'm not sure you could detect water molecules on the surface of wood, but learning a bit more about your goals and constraints may help! Thanks!

Is this a question? Click here to post it to the Questions page.

Reply to this comment...

Frequently, in agriculture, near infra-red (NIR) is used, often for grain ananysis. One item is moisture. They will change the amount the farmer is paid based on the moisture in the grain. This is often determined by NIR.

So the first place you should start is with NIR.

Now NIR can get pretty nasty with reflectance, then first derivative, and second derivative plots. So be aware.

Reply to this comment...

Wood moisture content is commonly measured at different points in the process of turning trees into finished wood products. Many wood moisture meters measure the electrical conductivity between two pins pushed into the surface of the wood. Others measure the magnetic field.

Reflected light might tell you something about the surface of the wood, but not much about the moisture content of the wood under the surface.


Reply to this comment...

There are a number of publications in this area because of the increased use of renewables (e.g. wood) for heating and the increased fire danger that is resulting from climate change. A couple of quick points from the background research I did on moisture content of wood (note that this relates to woods and not to grains or grasses).

  • it is necessary to penetrate the surface of the wood by1/5 to 1/4 of it's thickness to get a good measurement of moisture content. The moisture content can vary quickly and dramatically from the surface through a section of wood.
  • different woods will provide different surface and subsurface moisture content readings

The most common techniques to use are either weight or conductivity.

From personal experience with security systems, measuring the surface characteristics to understand the moisture content presents some technical issues. For example, under common environmental conditions chain link fence can be found to have a high moisture content.

It is certainly a challenging and interesting question.

Reply to this comment...

Hi @MarcioPamplona, Step one is to figure out where the absorbance shows up. By that I mean the wavelength that you need to detect. Googling "wood moisture nir" gave a link to an article in the Journal of Near Infrared Spectroscopy (2000, vol 8, p183). While I could not get the text of the article, the abstract indicates that bands are observed around 1450 and 1930 nm. Question 2 is: what kind of detector will you need to observe this wavelength? Googling "near infrared detectors" got this: The pdf seems to suggest that silicon based ccd's are useful to 1000 nm and above that, you need a specialized detector (ie, silicon based ccds like those in webcams will not work). ThorLabs will sell you a detector that should work for about $200 ( That will give you one pixel. Of course, you will still need to build something that will separate wavelengths for you and allow you to make background and reference measurements. You can probably use some kind of chopper with filters to do this, so I think you can get away without a grating. Still, this is a pretty daunting project. Good luck. Jack

Is this a question? Click here to post it to the Questions page.

That's a lot of money.
Nir is using overtones of IR bands. So you might be able to get by with something like atr with an IR. Just can't see that saving money. Most irs run more than $10k. That's before the atr accessory. And the atr only covers a very small area. Hmmmm.

Reply to this comment...

Hi, all - i believe I've fully recovered this post -- apologies for the database problem over the weekend. Please tell me if things aren't working normally here since the fix.

Reply to this comment...

Yes, i'm aware of the technologies. I want to develop a sensor, using NIR spectometry to measure the moisture level and characterize wood chips.

Like this comercial sensor: 0.1% to 0.5% of accuracy.

How to achieve some similar result and where i start ?

Is this a question? Click here to post it to the Questions page.

Reply to this comment...

Call them on the phone and talk to their salesman. Talk to tech services of you have to, but sales is usually better. A word of warning, the sales guy will want to know what impact it will have on him. The more impact ($), the more helpful. Do the same with any competitors. They can tell you weak points.

Then we can talk more about sendors.

Reply to this comment...

Log in to comment