14 C O N T E X T 1 6 4 : M A Y 2 0 2 0 laboratory can give good, semi-quantitative estimates of moisture levels. However, some devices perform better than others. Of all the capacitance and resistivity meters tested,the CEM capacitance and the Resipod resistance devices gave the best estimates of moisture content over a wide range of moisture condi- tions, from near-dry to near-saturated. Other proprietary conductivity and capaci- tance meters gave a variable performance. This was not surprising given that they are calibrated for a particular material. For exam- ple, Protimeters are precalibrated for timber, so they will be unreliable when used quantitatively for stone, brick or plasters. Furthermore, the majority gave consistent results only over a narrow range of moisture contents and were unreliable at near-saturated conditions. All were affected by salts and metallic elements close to the surface of the measurement loca- tion. Consequently, these types of devices are best used for comparative analysis in the low-to-mid range of the hygroscopic region, and where there are no salts or metals near the surface to interfere with the readings. Their main advantage is that they can provide a rapid, non-destructive method of identifying levels of moisture by giving comparative measurements. Performance also varied depending on the type of building material. Comparing three electrical meters (capacitance CEM and FMW and Protimeter in M-mode), the CEM meter gave more consistent readings than the other devices over a wide range of moisture content for limestone and brick. The Protimeter and the FMW meter only produced reasonably reliable readings for lime- stone and brick at moisture contents roughly below three and four per cent respectively. Another disadvantage of these meters is that they are only capable of one-off readings. In most instances for historic buildings monitor- ing of long-term trends is necessary as condi- tions are never static. Evaluation of changes of moisture contents can be used in different contexts, from monitor- ing the effect of an intervention, to drying after a catastrophic event, to assessing seasonal dif- ferences. The readings provide proxy moisture contents which are indicators of condition only. Monitoring will identify whether the material is drying, getting wet or has reached equilibrium. There are several ways of monitoring in situ, of which resistivity and humidity measurements are the easiest to implement. Both techniques can be resource heavy as they require a power supply, data logging and sensing equipment, but once installed they can be managed remotely. The sensors can be attached to the surface, embedded or fixed at the interface between the wall and another material. In the field, resistiv- ity (wooden blocks attached to electrodes) and Honeywell humidity sensors placed at the inter- faces between fabric and internal wall insula- tions have proved to be successful. However, initial laboratory testing of an embedded Rotronic humidity sensor only provided wet or dry readings, suggesting that humidity sensors may not recover if kept for prolonged periods at relative humidities greater than 98 per cent. At saturated conditions, the same could apply to the timber resistance sensors. Given that the area of interest is the hygroscopic region, argu- ably this is an acceptable level of risk, and field experience has shown that both types of sensors do recover as moisture levels fall. There are many factors that affect the assess- ment of moisture. Interpretation is challenging as the techniques are based on a variety of different principles. Many only have arbitrary scales that display values from 0-100, 0-200 or into the thousands. These values are simply non-absolute indicators of higher or lower moisture content. Furthermore, the value of the measurements depends on the experience of the user to interpret the results. Practical issues of handling between practitioners also influence the readings: some require good contact, and with others the amount of pressure applied or the degree to which the probes have penetrated a substrate affect the measurements. Calibration is another factor as commercial devices are usually calibrated for one type of material. Manufacturers sometimes pro- vide conversion tables or specific calibration values for generic materials, but evaluating their accuracy is difficult because of the non- homogeneity of historic materials, ageing, salts and weathering. Most historic buildings are composite structures, which adds another level of complexity. Moisture map of a section of a wall using the microwave moisture meter: (clockwise from top) spreadsheet with measured values; digital image from the meter; visible light image Note The experiments described in this article on evaluating a range of moisture measurement methods are for specific applications for historic building materials.These applications are not the standard uses for which the instruments were designed. Any comments or results given about their performance are specific to the tests conducted and do not represent any criticism about their application under standard use. Acknowledgements HeatherViles, professor of biogeomorphology and heritage conservation, and Hong Zhang, research technician, both at the school of geography and the environment, Oxford University; Scott Orr, lecturer, UCL Institute for Sustainable Heritage; Iain McCaig, senior building conservation advisor, Historic England Soki Rhee-Duverne is a building conservation advisor for the building conservation and geospatial survey team at Historic England, specialising in energy efficiency and the hygrothermal performance of historic buildings.