I was exposed to excessive levels of ultraviolet radiationApril 13, 2011
“The spectral emission of xenon lamps, which at wavelengths shorter than infrared closely matches that of a black body radiator at about 6000 K, enables their use as solar radiation simulators. Their emission spectrum is continuous from the UV through to the IR regions. Large amounts of UVA, UVB and UVC are emitted by unfiltered lamps to the extent that they can present a significant health hazard if incorrectly used.”
Environmental Health Criteria 160: Ultraviolet Radiation
During the years from 1993 to 2010 I and several coworkers were unknowingly exposed to very high dosages of UV radiation. Our hands especially may have been exposed on a daily basis to levels that were many times the level that is considered “safe”. On some occasions we may have been exposed to levels that were 100 times the “safe” level. This was due to a piece of code used to calculate our exposure that had an inverted unit conversion which caused one of its parameters to be off by a factor of 10,000.
Excessive UV radiation can have health consequences. Sunburn is a simple and obvious example of this. Less obvious is the correlation between years of UV exposure from the sun and some malignant and nonmalignant skin cancers and cataracts.
Health professionals urge people who spend a lot of time in the sun to take certain precautions, such as sun screens and hats that provide good shading to the face and neck. It turns out that our workhorse solar simulator had far more UV radiation on its work stage than, say, the sun at noon in July. A whole lot more.
The UV danger for a light source is quantified by a “Threshold Limit Value” or TLV. In the case of ultraviolet radiation the TLV defines the number of seconds in a day that a person can be exposed without causing harm. The American Conference of Governmental and Industrial Hygienists (ACGIH) provides a formula that will translate the spectrum of a light source into a UV TLV for that source.
How realistic is the ACGIH formula for calculating the TLV?
We can apply the ACGIH formula to standard solar reference spectra to get an idea. Consider these three standard spectra: AM0, AM1.5 global and AM1.5 direct. AM0 is the solar spectrum at the top of the atmosphere. AM1.5 direct represents the spectrum on the ground from the rays that come directly form the sun and cast a hard shadow. AM1.5 global represents the direct rays plus the light that is scattered to the ground by the rest of the sky. These are reference spectra – your spectrum will vary by location, season, time of day, weather, etc.
The ACGIH formula gives 0, 1000, and 1000 seconds as the threshold limit values for the AM0, AM1.5 direct and AM1.5 global reference spectra respectively. Luckily for us, the atmosphere removes most of the UV from sunlight before it reaches the ground. But if you are in space you better wear a spacesuit or use very good sunscreen.
Is 1000 seconds (about 17 minutes) a realistic TLV for AM1.5 global or direct? In other words, will the top of exposed shoulders or a bald head suffer UV damage after about 17 minutes at noon on a sunny day in July in the central US? My gut, unscientific, answer is that most people would not suffer a problem after 17 minutes. But after an hour many fair-skinned people would be effected. After two hours most fair-skinned people and many dark-skinned people would be effected. Years of daily exposure may take a serious toll.
What happened at NREL?
In 1993 the optical filtering of our primary one-sun solar simulator was modified. The spectrum was measured and code written at NREL to implement the ACGIH formula was used to calculate the UV threshold limit value. One of the parameters of the ACGIH formula, the “effective irradiance,” was presented in the ACGIH literature in units of microwatts per centimeter squared (μW/cm2). The author of the code preferred units of Watts per meter squared (W/m2).
To convert from μW/cm2 to W/m2 you must multiply the orignal value 0.01. The code author got it backwards and multiplied by 100. Thus the conversion was in error by a factor of 10,000 (1002).
When the code was applied to the new simulator spectrum in 1993, it reported that the threshold limit value for exposure to the solar simulator light was 1000 seconds. Although the simulator spectrum was regularly measured during the following 17 years, the conversion error in the code was not discovered until 2010. It turns out that a proper implementation of the code would have yielded a threshold limit value of only 2.6 seconds for the 1993 simulator spectrum. Roughly speaking, we would receive the same harmful effect to exposed skin from the UV in our simulator in 2.6 seconds that a person walking outside on a sunny summer day would receive from the sun in 17 minutes.
Other spectral measurements of the solar simulator taken between 1993 and 2010 show the TLV value coming in at about 13 seconds. These very short TLV values were discovered by me in mid-2010, and subsequently confirmed by an outside consultant.
My best guess about the TLV
The NREL solar simulator in question is used to simulate the sun for testing photovoltaic devices. The SOP (Safe Operating Procedure) stated that the TLV value for exposure was 1000 seconds, and the operators (including me) did not think there was a problem. But the simulator spectrum had a bump in the UV that forced the threshold limit value for exposure to a time much shorter than for the actual sun.
Several operators used this solar simulator on a daily basis for years. They must have been exposed to many times the TLV for UV radiation on a regular basis. At a TLV of only 2.6 seconds it would take only about 4 minutes to be 100 times over the TLV. Amazingly, no injuries were ever reported. How can that be?
Long term effects, such as skin cancer, could take years to develop. Nevertheless, nobody ever even reported an acute “sunburn.”
My guess is that the actual health effects of different wavelength ranges of ultraviolet light are poorly understood, and the wavelength weighting applied by the ACGIH formula is as much excessive caution as it is science. The human health effects of UL light with wavelenghts shorter than about 300 nm (where the solar irradiance at the surface of the earth at low elevations goes to nearly zero) are probably weighted very highly in the calculation of the TLV. This covers doubt with a thick safety cushion.