Youve made an offer on a used boat and have retained a marine surveyor to examine the vessel prior to completing the transaction. One of the more important things he or she will do is make some sort of assessment about the condition of the hull. Delamination of the skins from the core or osmotic blisters are serious problems and may cause you to nix the deal.
Over time, all fiberglass hulls absorb moisture. Small amounts are not of great concern. Even higher amounts may not be an indication that the hull has or ever will blister, of which there are various causes.
Still, the prudent buyer and surveyor will want some indication of a hulls level of wetness. To determine this, the most common tool is the moisture meter. These small, handheld devices, generally costing less than $500, were originally developed for other industries, such as timber, roofing and building trades. They were adopted by the marine industry, but in many cases the makers of these instruments have made no real effort to modify them for use on fiberglass. Hence, the readings one might get from a fiberglass hull must be extrapolated from a scale designed for wood or concrete. This is your first clue that the moisture meter is an imperfect device for determining moisture levels in a fiberglass boats hull.
At Practical Sailor, we receive quite a few calls from readers who have questions about how best to diagnose and repair a blistered hull. Often, they tell us that the yard or surveyor used a moisture meter and found unacceptably high levels of moisture. Occasionally, the owner is advised to have the gelcoat peeled, perhaps even some of the laminate, then replaced. Our advice to them is to first, before committing to such an expensive repair job, try determining why the hull has blistered. If you don’t know the cause, you can’t prescribe the cure. Simply popping blisters and filling them with epoxy putty wont stabilize a hull laminate that still has uncatalyzed hardener in it. And in some instances, blisters are not indicative of deteriorating laminates. For example, solvents trapped during the application of bottom paint can cause blisters that may look like trouble, but are in fact harmless.
Before undertaking any major repairs, retain an independent surveyor, rather than rely on your boat yards recommendation. After all, its the yard which stands to make a profit on your repairs.
Many laminate experts advocate cutting a hull plug with a hole saw and having it analyzed in a laboratory. The procedure is destructive, to be sure, but the $300 or so cost of this test could save you thousands if it is found that the laminate is basically sound in spite of high moisture meter readings.
But not all experts agree, noting that the common lab tests to determine percentage of moisture and detection of gases still fail to pinpoint the exact cause of blistering, or predict the likelihood of blistering.
How Much Do Surveyors Know?
Last year, we collaborated with marine surveyor, Jonathan Klopman of Marblehead, Massachusetts, in the testing of five leading brands of moisture meters-Sovereign, Tramex Skipper, Caisson Novanex, Protimeter Aquant and Protimeter Surveymaster SM. Prior to these tests, Klopman, sponsored by Professional Boatbuilder magazine, also distributed a questionnaire to 94 fellow surveyors. The results are eye-opening.
Ninety percent said they owned either the Sovereign or Tramex. Forty percent have been using the instrument for one to five years. The majority use it weekly. But when asked if they understood the operating principle, 33% answered incorrectly and 22% said they didnt know!
Fortunately, most seem to understand their meters limitations. While 60% said they trusted the readings most of the time, this same number also admitted that their meter could not provide an accurate prediction of an osmotic problem when there are no blisters evident on the hull .
Regarding how deep into the laminate the meter can read, 63% said they werent sure. After testing them, we now know why they werent sure: Depth varies by brand, scale settings used, and the meters may be fooled by additives in bottom coatings or the laminate itself.
Nevertheless, surveyors use moisture meters more than anybody else and throughtraining programs conducted by organizations such as NAMS (National Association of Marine Surveyors) and SAMS (Society of Accredited Marine Surveyors), their level of knowledge is increasing.
How They Work
Most moisture meters are of the capacitance type. These meters have two electrodes or sensors, one that transmits an AC signal and another that receives it. Because water has a much higher dielectric constant or permitivity than air or fiberglass, the difference in strength between the transmitted and received signal can be measured and displayed on the meter, either by an analog needle or digitally in an LCD window.
Some meters, like the Sovereign, also have two pin probes for measuring moisture in wood; these work by resistance and use a DC signal.
A third method, used by the Protimeter Aquant and called the free-field effect, is slightly different. Instead of two electrodes next to each other, the Aquant has a transmitter in the head of the instrument, which directs a signal into the laminate where it is redirected by any moisture present back to the second electrode in the handle. In effect, the water acts as an aerial to re-radiate the RF signal.
A number of fiberglass panels were laid-up for testing with the meters. These consisted of alternating layers of mat and woven roving with general purpose polyester resin, ranging in thickness from .135″ to .672″ (to achieve the 1″ thickness shown in the graph above, we combined several panels). Additional panels were made up of Coremat.
We also made up some epoxy panels with graphite, aluminum and copper powder, barrier coat additives and carbon fiber.
Additional tests were performed to see if frozen water gives different readings, and to see if different types of bottom paint affect readings.
Lastly, we tested to see if battery voltage makes a difference in the readings. All of the units use 9-volt batteries, except the Protimeter Surveymaster, which uses two AA batteries.
All of the above tests were performed indoors on a bench. For the thickness tests, the panels were placed on top of two pads of wet carpet, soaking in a shallow tray. Reference readings were taken of each panel before setting them on the wet carpet. A second set of tests used paper wetted alternately with fresh and saltwater. In the accompanying graphs, these reference numbers are subtracted from the wet readings to show the actual moisture reading. We started with the thinnest panels, progressing to thicker panels in fine increments so that we could detect when the meter stopped penetrating the laminate all the way to the water.
After two series of bench tests, the first conducted by one of our contributing editors, a nuclear physicist, and a second series conducted by Klopman and PS staff, we took them to several boat yards and tried all five meters on a variety of hulls, some known to be wet, some blistered, and some dry.
Though all of the meters have graduated scales, some indicating percentage of moisture, these are not true representations of the amount of moisture in a fiberglass panel (in part, because most of these meters are not designed for fiberglass). Consequently, we recorded our readings as a percentage of full scale.
Further to the matter of moisture percentage, a fully saturated fiberglass hull will have no more than 2.5% to 3.5% moisture. Most of the meters designed for timber sound an alarm at 20-25% moisture, which is a typical fiber saturation point for many species. Some meter makers provide a conversion card for timber to fiberglass (eg., 25% wood = 3% glass), but few surveyors put a lot of stock in it. Hence there is no correlation between the percentage of moisture indicated by the meter and the actual amount of moisture in the laminate.
What We Found
While there were some anomalies that are difficult to explain, and some variation in the responses, due perhaps to differing conditions (e.g., how the meter head is oriented to a surface can have a significant effect on the readings), we did arrive at some basic conclusions.
Battery Test. All of the meters automatically compensated for low battery voltage except the Sovereign, which must be continually set to zero to account for changes in battery voltage. This is done easily with a knob and the needle, but it is not necessary with the others. Also, the Protimeter Aquant will not turn on below 8-volts, which one presumes is near its minimum operating voltage.
Thickness Test: As expected, the thicker the laminate, the more difficulty the meters had in detecting water, even though the same amount of water was present underneath each panel. If a meter had more than one scale for sensitivity, we tried all of them.
With all meters, the readings would be quite low with a thick laminate, which could lead an inexperienced user to conclude that there isn’t much moisture present. This is especially confusing where the meters scale incorporates colored zones-green (good or low), yellow (fair or moderate) and red (bad or high), or colored lights, as in the case of the Protimeter Aquant and Surveymaster. What this means is that a moisture meter cannot tell you where the water is in the laminate or precisely how much. A lot of moisture deep in the laminate may give the same reading as a small amount of moisture near the surface. Knowing the thickness of a hull laminate ahead of testing is helpful, but in many boats wont be known. Plus, thickness varies throughout the laminate, from topsides to bottom, and even within the same general area.
The order in which the five meters ranked, from deepest to shallowest, was: Tramex, Protimeter Aquant, Protimeter Surveymaster SM, Caisson Novanex V1-D3 and Sovereign. (See table on page 5.)
Frozen water. We ran two different tests to determine whether the meters could detect frozen water. In the first, we compared a sample of .135″ laminate and wet paper frozen. No meter gave a reading differing from laminate alone. This suggested that the meters do not detect frozen water.
In the second set of tests, we took four panels in various stages of wetness and deterioration: moist, saturated, advanced brown rot and white rot. Reading were taken wet. Then the panels were frozen and tested again. Comparing the wet and frozen readings, most of the frozen readings were lower than the wet, but not all. We cannot account for those few frozen readings that were the same as room temperature readings or higher.
While the tests werent the same, the discrepancies leave questions. Still, the bottom line is not to trust any moisture meter readings taken when the laminate is below 32F. Wait until the temperature warms and the laminate has thawed.
As a side note, glycol, a byproduct of the resins used in boats and the cause of some blistering, will not freeze and can be detected in sub-freezing weather.
Polar compounds. A thin laminate with raw catalyst trapped in the layers was measured. This showed no significant difference to the normal panels with fully reacted catalyst.
Isopropyl alcohol, epoxy and hardener were placed in small, separate divots in a panel and covered with a poly sheet. This is a very crude and non-quantitative test that only checks for gross sensitivity. The Tramex and Protimeter Aquant showed no sensitivity to the compounds. The Sovereign and Caisson did. The Protimeter Surveymaster SM was not in hand during this test and so was not tested.
Metal detection. When we placed a sheet of aluminum foil under a panel, all of the meters gave readings similar to wet paper.
Next, we mixed copper, aluminum and aluminum-based WEST System 422 barrier coat additive into epoxy resin and brushed it on a bare Masonite panel. We tested the panels both dry and with wet paper behind them. The meters did not react significantly differently to any of these additives. Dry, the panels with additives read about the same as the bare Masonite panel. The same was true with wet paper behind.
When we performed the same test with carbon fiber, however, every meter was pegged or nearly so. This means that any hull with carbon fiber cannot be tested for wetness with a moisture meter.
Lastly, we performed the same test on graphite powder, because it is sometimes used as a pigment in black bottom paint. Here we noted an increase in the readings of some meters. For example, the Tramex read about 5% (relative scale of 1-100%) on the bare panel and about 55% with graphite. The Caisson read 7% on the bare and about 17% with graphite. The Protimeter Aquant, because its signal appears to begin reading just beneath the surface (an advantage in avoiding false positives owing to surface moisture) did not react significantly to the graphite-epoxy coating. The bottom line is that any unusual reading taken on black bottom paint should be investigated for the presence of graphite. Safer is the common recommendation to scrape away bottom paint and place the meter head directly on clean gelcoat. Be wary, however, of paint remaining in the gelcoat pores.
Owners should also know that these meters will often detect the steel webbing inside a rudder, sometimes even the copper foil used as a counterpoise for SSB radios, and condensation on the inside of the hull. But the meter doesn’t know the difference between metal and moisture, whether its inside the laminate or harmless condensation. Its alarm just beeps.
In addition to bench and field tests, we also evaluated each meter for quality of construction, handling, visibility and overall ease of use. Following are our comments, plus a brief description of each meter.
The UK-made Sovereign comes in a heavy-duty binocular-type case. Unlike most of the others, it comes with instructions, but no mention is made of fiberglass. The separate transducer head is connected via a cord. There are both probes for wood and a disc-shaped head for hard surfaces.
Heavy duty metal case and dials.
Instructions for calibration and use.
Battery check (but no warning) and calibration.
Remote head is compact, but must be unplugged to fit in case.
Relative scale included (0-100%).
Adjustable alarm and zero control.
Plastic protective cover for head.
Easy battery access.
Pins for probing wood.
No power on light; easy to drain batteries.
Two-handed use is awkward.
Neither of its two scales are calibrated for fiberglass.
Scales not evenly graduated.
Pressure on head affects readings.
Doesnt compensate for low voltage.
Made in Ireland, the Tramex has an analog display with three scales. The transducer consists of two large pads on the backside of the same plastic case used by the Caisson Novanex.
Battery on light, but no check function.
Scale evenly graduated and easy to interpret.
Relative scale included (0-100%).
Scale switch inexpensive.
Three scales not sequential from sensitive to less-sensitive (i.e. #2 scale is most sensitive, #3 least sensitive).
Case appears fragile.
No automatic off feature.
Small display difficult to read.
Made in the UK, the Novanex has a black plastic case. The display is digital. Like the Aquant, the head is at the forward end of the case.
Backlit LCD display easy to read.
Small footprint; easy to orient head.
Hard plastic head appears durable.
Adjustable zero, alarm threshold, upper end of range (0-15).
Low battery indicator.
Case appears fragile.
No audio alarm.
No automatic off feature.
Made in the UK, this meter differs from the others in that it transmits a radio frequency (RF) which is returned to the receiving electrode through the body of the user. Moisture in the laminate acts as an aerial that re-radiates the RF to the lower half of the instrument. This is called a free field effect. A transfer electrode or near field shield may be placed over the transmitter head for shallow readings. Placing a hand on the surface being measured affects the readings.
The transmitter head is along the forward end of the case and has a small footprint. To take readings, a single button is pushed. Readings are displayed by lighting up individual LED lights.
Easy to see readings, except in direct sunlight.
Will not turn on at low voltage (8V).
Space to carry spare battery.
Automatic off (but time interval too short).
Adjustable alarm threshold.
Switch not sealed.
Short, coarse scale (3 flickering lights = +/-15%).
No battery check feature.
Case appears fragile.
Head subject to abrasion.
Must be held 90 to surface, which is sometimes difficult.
Near field attachment unreliable at thicknesses over 0.125″.
Protimeter Surveymaster SM
A relatively new product, the Surveymaster has a display similar to the Aquant, but the scale is different. Also, the transducer head is on the backside of the meter, covered with plastic, but easy to rock. The remote head and separate probes are attractive features. When using the remote head, the reading is displayed in a small LCD window.
Durable rubber case gasket.
Remote head and probes.
Easy battery access; good battery contacts.
Automatic off (but time interval too short).
Adjustable audio alarm.
Head difficult to keep at 90 to surface.
Because of head location on back, holding hand tends to obscure the LCD lights.
No numbers on scale.
The moisture meter is actually a fairly simple device. Some have called them glorified stud finders. A moisture meter cannot tell you the soundness of your hull laminate. But it can give you worthwhile clues on which to pursue further investigation.
In use, meter readings first should be taken in areas believed to be dry or at least typical of above water laminate. So, before checking the bottom, take readings in several areas of the topsides to obtain a general base number. Then begin testing the bottom. If the boat has just been hauled, allow some time for surface moisture to evaporate. It is best to scrape away the bottom paint and place the meters head directly on clean gelcoat.
Readings that vary significantly from the base may be worth investigating, certainly if there are blisters present. But if no blisters are present, even high readings may not correctly indicate problems in the laminate. When this is the case, you may elect to do nothing, at least for the time being. If the boat has been hauled for the winter, check again in the spring. If readings are still high (and they probably wont change a great deal if the laminate is extremely wet because it is difficult for the moisture to evaporate when entrapped), then you might consider lab analysis of a core sample or grinding down the gelcoat to reveal the inner plies of the laminate.
From the above, one can see why surveyors often rely more on their meters to measure dryness of a hull rather than its wetness. They don’t need a meter to detect surface blisters, but they do need help knowing when a hull is dry enough to recoat.
Moisture meters are also helpful in determining whether through-hull or deck fitting bedding has failed. If you get unusually high readings around, say, a deck cleat, it is advisable to remove the cleat and rebed it. If the bolts are through balsa core that is dripping wet, this may be the time to take the more radical step of cutting away the bad balsa and filling each area with an epoxy slurry.
Similarly, a meter may be helpful in determining the extent of wet coring in a deck.
After spending many hours with each of the five meters, we concluded that we prefer the Tramex for its superior ability to read through thick laminates. Had we the funds, wed also buy the Protimeter Aquant for its ability to ignore surface moisture. The near-field shield gives it some flexibility the others lack. The Sovereign, long the most popular meter, does not penetrate deeply through laminates, is awkward to use and is the most expensive of the group. The Caisson Novanex, which has the same case as the Tramex, is OK, but not our first choice. The Protimeter Surveymaster SM does not have a printed scale, forcing one to count lights or take special note of their colors. Because its transmitting electrode is on the backside, it is easy to cover the lights with your hand.
Contacts- Novanex, Caisson Audio Products, Toernooieveld 116, 6525 EC Nijmegen, Holland. Tramex, Tramex Ltd., Shankill Bus. Ctr., Co. Dublin, Ireland. Sovereign Chemical Industries Ltd., Barrow-in-Furness, Cumbria LA14 4QU England. Protimeter Aquant and Surveymaster, Protimeter North America, PO Box 450, Danvers, MA 01923; 781/938-3966. Novanex, Caisson Novanex and Sovereign available from US distributor: J.R. Overseas, PO Box 370, Kent, CT 06757; 860/927-3808, e-mail firstname.lastname@example.org