Q: We purchase hardwood dowels with a moisture specification of 7.0 to 9.0 percent MC. We see cracking if the dowels are a little wetter, and we do see some coming into us at 10 to 12 percent MC. Is it unreasonable for us to insist on our tighter MC specification? We are rejecting pieces outside the limits.

A: With any MC specification, you need to indicate how you will be measuring the MC. Sometimes, the meter itself will put in substantial variation in the values. The oven test would be the best, but it might be impractical. Therefore, as you check the MC, take some of the wettest and some of the driest pieces and double check the MC using the oven test. This test will give you confidence (or maybe lack of confidence) in your meter. Make sure that your supplier also agrees with how you will measure the MC and that this technique and brand of meter can be used to reject wet or dry pieces.

Now, let's consider your question about moisture variability. When hardwood lumber is dried, a good operation will attain a standard deviation of moisture content variability of 0.6. A very good operation will achieve 0.5. With a little effort, variations can be even less. A wide spread of MC is preventable or avoidable, if desired. For softwoods, many operations are not using moisture samples, but dry on the basis of time alone. As a result, some softwood drying operations have exceptionally variable MCs and do not know how to improve.

What is the standard deviation (SD)? It is a measure of variability. It can be calculated easily on a \$20 calculator. In fact, many moisture meters have this calculation done internally for you. Here is the neat thing about SD. If you sample 30 pieces randomly in a load of lumber and calculate the average MC and the SD, the SD, when added and subtracted from the average MC, will include about two-thirds of the MC values; when two SD are added and subtracted, the interval includes 95 percent of all the MC values; and three times SD includes 99.9 percent of all the values. (Example: If the average is 7.0 percent MC, and the SD is 0.5, then the range of 6.5 to 7.5 percent MC will include two-thirds of the MC values; 6.0 to 8.0 percent MC will include 95 percent of the MC values; and 5.5 percent to 8.5 percent MC, will include virtually all the values.)

Is your specification (which is equivalent to SD = 0.33) unreasonable? Yes, perhaps slightly so from a kiln drying perspective where speed of drying is important and longer drying times mean less profits for the kiln. You might find a better specification is that you should allow 2-1/2 percent of the pieces to be higher and 2-1/2 percent to be slightly lower (this is SD = 0.5) than the + or - 1 percent MC range, if you want typical drying. However, if you are willing to pay slightly more for the drying (perhaps \$25 per MBF for one more day of kiln drying), the SD can be substantially tightened.

On the other hand, from a manufacturing standpoint, your specification is totally reasonable, and perhaps even essential for a high quality product. Indeed, with an expensive product (especially when the product fails after you have put some labor into it), you can certainly afford better drying. I would insist on such tight MC values, which is a SD = 0.33. If the kiln cannot achieve these values, then reject the pieces outside the limits. Then, store them in a warm, controlled RH condition (100 F and 30 percent RH) for a few weeks, allowing the MC to adjust, and then check them again. Of course, if they dried a little they may shrink and be too small. They also could warp slightly.

Summary: Have the kiln operation do the best it can (SD = 0.33), pay a little extra, and then check quality yourself before you use them.

Have something to say? Share your thoughts with us in the comments below.