Electrical conductivity of supercooled cloud water at Mt. Washington, NH

Aircraft icing, despite improved forecasts and onboard ice detection and protection systems, still causes crashes and loss of life. As a result, research is being conducted to develop improved methods of removing accreted ice despite the proven record of deicing boots and heated leading edges. One of these new technologies, developed at Dartmouth College by V. Petrenko, utilizes a grid of cathodes and anodes on the leading edge of a wing to conduct a direct current through accreted rime or clear ice. The current causes electrolysis, decomposing the ice into its gaseous oxygen and hydrogen components. This causes the ice to erode at the ice-substrate interface, debonding, and removal by the slip stream. Electrolysis debonding of ice is most efficient when ice has a high electrical conductivity. Low conductivities require large currents which can damage the anode-cathode grid through arcing. Cloud water electrical conductivities vary depending upon geographic location and air mass source region. An electrolysis-based deicing system would need to be functional in all icing situations, and should function in the lowest water conductivity conditions that the aircraft is likely to encounter.

The objectives of this report are to characterize cloud supercooled liquid water electrical conductivity and chemistry measured at the summit of Mt. Washington, NH, to determine its potential variability in the free atmosphere. During the winters of 2001 and 2002, cloud water electrical conductivity and chemical composition were measured, and source regions of air masses were back cast to assess causes of conductivity magnitudes. Measurements made during the winters of 1988 and 1989 during another program augmented the 2001-2002 measurements.

Rime ice was sampled on 2.5-cm diameter by 38-cm long acrylic rods exposed at Mt. Washington Observatory (MWO) until about 1-cm of clear ice or about 2-cm of rime ice had accumulated. Clean room techniques were used, as possible, in winds as high as 30 m s-1 to expose and to remove the rods. Rods were exposed by MWO staff trained in chemical sampling, and conductivity and chemical analyses of melted ice were performed by the University of New Hampshire. CRREL performed subsequent trajectory analyses and summaries.

Sampling during the winters of 1988 and 1989 on Plexiglass rotating multicylinders yielded 53 samples with a minimum conductivity of 1.2 µS cm-1, a maximum of 124 µS cm-1, and a mean of 24.2 µS cm-1. Sampling during the winters of 2001 and 2002 yielded 147 samples with a minimum conductivity of 0.6 µS cm-1, a maximum of 443 µS cm-1, and a mean of 56.5 µS cm-1. In experiments conducted at CRREL, ice created from water of 10.1 µS cm-1 was not sufficiently conductive to support electrolysis. Petrenko has not published a minimum conductivity for utility of the electrolysis method. However, taking 10.1 µS cm-1 as a minimum acceptable conductivity, 43% of the 1988-1989 samples and 8% of the 2001-2002 samples were less conductive.

We also assessed trajectories of air masses carrying the liquid water samples to Mt. Washington, and analyzed the time period that each air mass might be exposed to point source locations of emissions of over 1000 tons year-1 in the United States and Canada. HYSLPIT_4 was used to calculate 96-hour trajectories using information from NOAA’s Eta Data Assimilation System.

In general, conductivity is lowest when air masses originate from the Atlantic Ocean, and highest when originating from the populated Middle Atlantic, Midwest, and southern Canada areas. Since easterly winds from the Atlantic Ocean are least frequent at Mt. Washington, then low conductivities are infrequent - as the conductivity statistics also show. However, even a low percentage of failure due to low conductivities may not be acceptable for protecting aircraft from dangerous icing conditions.