Measuring moisture content: oven-drying and meters

One reference, several tools

The oven-dry method follows the definition of moisture content directly and is the reference. Meters estimate the same quantity from an electrical property and need species and temperature corrections.

Oven-drying: the reference method

According to Bergman (ASTM D4442 Method A, North America):

• Cut the specimen at least 500 mm from the board end, free of knots, bark and pitch pockets. A lumber specimen is a full cross-section about 25 mm long.
• Weigh it at once.
• Dry it at 101 to 105 °C and re-weigh until the mass changes by no appreciable amount over 3-hour intervals; typically 12 to 48 hours for a 25 mm section.
• Avoid microwave ovens; the end point is unreliable (Conners).

The method is accurate but slow and destructive, and can slightly overestimate moisture in woods with volatile extractives, because lost compounds are counted as water. Its role is to calibrate everything else and to settle disputes.

Worked example 1: a slice from a board

A 25 mm full-section slice weighs 86.4 g when cut and 64.0 g after drying to constant mass.

• Water: 86.4 - 64.0 = 22.4 g.
• MC = 22.4 / 64.0 x 100 = 35.0 %.

This board is above the FSP, so a pin meter should not be trusted on it (see below), and the oven-dry result is the right way to find its moisture content.

Resistance (pin) meters

A resistance meter drives two pins into the wood, measures electrical resistance between them and converts it to moisture content with a calibration curve.

• Range: about 7 to 30 % (Boone and Wengert, University of Wisconsin-Extension); Bergman gives about 6 to 30 %. The sources differ at the margins, so check the instrument manual. Their shared message: above about 30 % the readings are not reliable, because resistance changes too little and varies too much.
• Accuracy: one meter manufacturer quoted by Conners gives +/-0.5 % at 6 to 12 % loosening to +/-2 % at 20 to 30 %. Used properly on lumber dried below the FSP, Bergman says electrical meters should give an average within about 1 % of the true average.
• Pin depth: pins driven to about one fifth to one quarter of the thickness give an average reading; shallow pins read the surface layer.
• Species: most US meters are calibrated for Douglas-fir or southern pine; corrections for other North American species are usually 1 % or less.
• Temperature: about 0.5 % MC per 10 °F of deviation from the calibration temperature (Bergman).

Worked example 2: pin depth and temperature

For a 60 mm thick board, one fifth is 12 mm and one quarter is 15 mm, so drive the pins 12 to 15 mm deep for an average reading. For a 50 mm board the range is 10 to 12.5 mm.

If boards are measured 40 °F away from the meter's calibration temperature, the temperature effect is about 0.5 x 40 / 10 = 2.0 points of moisture content. Check the instrument manual for the direction of the correction.

Capacitance and dielectric meters

These press a sensor plate on the surface and measure a dielectric property in the field the plate projects. They are fast and leave no holes, which suits incoming inspection and grading lines. Reported ranges are about 5 to 30 % (Boone and Wengert), 4 to 30 % (Bergman) and 5 to 25 % (Conners). Their key weakness is density: Boone and Wengert report that they are strongly affected by specific gravity, so use the species setting where available. A surface-sensing meter probably sees mainly the outer layer, so a wet core under a dry shell could go unnoticed (portal inference).

Which tool for which job

Key points

• Oven-drying at 101 to 105 °C to constant mass is the reference; it is slow and destructive.
• Electrical meters are reliable only below about 30 % MC and need species and temperature corrections.
• Pin depth of one fifth to one quarter of thickness gives an average reading.
• Dielectric meters are fast but affected by density and probably see mainly the surface layer.
• Meter accuracy ranges differ between sources; manufacturers' current specifications take precedence.