Summary and conclusions

Kortlægning og sundhedsmæssig vurdering af kviksølv i energisparepærer og lysstofrør

Compact fluorescent lamps contain small amounts of mercury which is hazardous to health

Energy-saving light bulbs (compact fluorescent lamps (CFLs)) represent one of the most efficient solutions which are available today to improve the energy efficiency for housing lightning – but compact and straight fluorescent lamps contain small amount of the element mercury which is hazardous to health. With this project, the Danish Environmental Protection Agency will examine whether a health risk arises when a compact or straight fluorescent lamp by accident breaks at home.

Therefore, it is examined

which types of compact and straight fluorescent lamps that exist on the Danish market for private use, and
which amounts of mercury and mercury compounds that are used in compact and straight fluorescent lamps on the Danish market for private use.

Based on this information, a risk assessment was made of a potential accident with a broken compact or straight fluorescent lamp emitting mercury vapour in a home.

The risk assessment has been carried out partly by use of a theoretical calculation of the amounts of mercury that is expected to vaporise, if a compact or straight fluorescent lamp breaks in a home; and partly by assessing measured concentrations in a home during weeks after an accident with a broken compact fluorescent lamp. The concentrations have been compared to known concentration values, where health effects have been observed.

This project has been carried out by FORCE Technology for the Danish Environmental Protection Agency during December 2009 to May 2010.

What is mercury?

Mercury is a metallic element that can occur as a metal, and in inorganic and metal-organic compounds. In addition, mercury is mixable with other metals – an amalgam is formed. Mercury (Hg⁰) is the only metal, which is liquid at normal pressure and temperature, and it appears as a heavy, odourless silver glistening liquid that has a relatively high vapour pressure at room temperature. Therefore, contact with liquid mercury will result in exposure to invisible and unnoticeable mercury vapours. Mercury vapours are seven times heavier than air and will spread along the floor in a room with inadequate ventilation.

In which form can mercury be found in light bulbs?

In compact and straight fluorescent lamps mercury is used either in the form of a HgFe compound, in the form of amalgam or in the form of metallic mercury. This mercury – either in a liquid or solid state – will be in a state of equilibrium with mercury in the vapour state. A small amount of mercury vapour will be present in the compact or straight fluorescent lamp, and it is because of this mercury in the vapour state, that the lamp will light up.

If one or more of the compact or straight fluorescent lamps are broken in a home, mercury vapour will be released to the indoor air, as long as the residues of the lamp have not been completely removed.

Consumers will be better protected against exposure to mercury in the lamps, if the lamps are using encapsulated mercury (in the form of a tablet or as amalgam) compared to the use of liquid mercury – as the mercury will be bound.

Why is mercury a health problem?

Especially mercury vapours are a health related problem, as a large part of the vapours are being absorbed through the lungs during inhalation. In contrast, the absorption of the poorly soluble metallic mercury through skin and through the gastrointestinal tract is minimal. Furthermore, mercury vapour can easily pass the blood-brain barrier and placenta, and can in that way have an impact on the central nervous system and the unborn child.

Most data on health effects of mercury vapour originate from occupational exposures. The lungs are the target organ at very high exposures to mercury vapours in the work environment. The effects are irritations and corrosions of the airways, and at a few hours of exposure to 1-3 mg Hg/m³ (1000-3000 µg Hg/m³) a deathly acute chemical pneumonia can occur. Immediate mortal danger can occur at exposure to a level of 10 mg Hg/m³ (10,000 µg Hg/m³).

The critical organ is the central nervous system at continuously high exposures to mercury vapours at levels of > 0.100 mg/m³ (>100 µg/m³). At these levels seriously damage as classical poisoning symptoms such as tremor, insomnia, depression, mental unbalance, irritability, memory loss, abnormal shyness and gingivitis is expected.

Light acute toxic effects (like hand tremors or memory loss) in humans can be anticipated at long-term exposure of 0.025-0,050 mg Hg/m³ (25-50 µg Hg/m³). This value of 0.025 mg Hg/m³ (25 µg Hg/m³) is identical to the Danish occupational threshold limit value.

Calculation of the level of mercury vapours with no health effects

Based on a health assessment of mercury vapour DNEL-values (Derived No Effect Level) of both short-term (30 minutes clean-up) and long-term (if a proper clean-up has not been carried out after an accident with a broken CFL) exposure has been calculated in this report. The calculated DNEL-values for short-term (DNEL_short value = 33 µg Hg/m³) and long-term (DNEL_long value = 0.4 µg Hg/m³) exposure are close to the Danish occupational threshold limit value (25 µg Hg/m³) and the American reference concentration (RfC) – a long-term concentration with no harmful effects (0.3 µg Hg/m³), respectively.

The LOAEL- (Lowest Observed Adverse Effect Level) and NOAEL-values (No Observed Adverse Effect Level) that are used for the calculation of the DNEL-values in this project, as well as the different threshold limit values that exist for mercury vapour, are all based on observations from the working environment of adult humans and their exposure to mercury vapours. As the documentation is based on occupational exposure, we do not know if children are more sensitive to exposure to mercury vapours. However, children are generally considered being more sensitive than adults towards toxic effects. Children have a lower respiration volume than adults and will therefore inhale less amounts of the toxic mercury vapours, but children also have a lower body weight than adults. Furthermore, their nervous system is under development – and as described in chapter four – the critical organ at continuously high exposure to mercury vapours is exactly the central nervous system. No assessment has been made in this project whether these concentrations will present a special risk to children that are present in a room where an accident with a CFL happens. The DNEL-values have, however, been calculated by use of a safety factor of 10 that accounts for the individual differences between humans. This safety factor should therefore account for the fact that children are more sensitive than adults towards toxic effects. However, information has been found, showing that exposure to mercury vapours is a special risk for pregnant, as mercury vapours can pass the placenta and harm the unborn child.

The calculated DNEL-values are then compared to the calculated worst-case concentrations during clean-up and to the concentrations measured after clean-up of broken compact or straight fluorescent lamps. These concentrations are found in different literature.

Which levels can be found on short-term exposure at accidents with broken bulbs at home?

For the calculation of the short-term exposure (the 30 minutes clean-up), a calculation model that accounts for ventilation in the room has been used. The calculation model uses a specific measure for the volume of air surrounding the exposed person – in this report called the breathing zone. This specific measure has been taken from ECHAs Guidance Chapter R.15 (2008), which describes that for short-term local exposure, the volume of air that immediately is surrounding the person can be set at 2 m³. The calculations show the following:

No ventilation: By using the assumption that 10 % of the mercury is evaporated during the first half an hour, the concentration in the room without ventilation will exceed the DNEL_short-value and thereby be able to present a health risk – no matter the amount of mercury in the CFLs. Calculated concentrations of mercury vapours (for 5 mg of Hg in a CFL) are exceeding the DNEL_short-value with up to 8 times. This is, however, an overestimation as ventilation of zero is a imaginary value as this corresponds to a theoretical air tight room with no ventilation through cracks.
Ordinary ventilation: slowly lowers the concentration of mercury vapours in the room. It will take up to two hours before the concentration of the mercury vapours in the breathing zone will be below the DNEL_short-value, if a “normal” ventilation rate occurs – and if only a single CFL is broken (for 5 mg Hg in a CFL).
Heavy ventilation: has a significant importance with respect to lowering the concentration of mercury to non-harmful levels at home after accidents. After 10 minutes of ventilation with all windows and door open, the mercury concentration in the breathing zone of 2 m³ will be below DNEL_short-value for CFLs with a low content of mercury (2.5 mg Hg), and will therefore not present an acute risk. After 15 minutes of heavy ventilation, the mercury concentration in the breathing zone is below the DNEL_short-value for CFLs with the allowed content today (5 mg Hg), and after 30 minutes the mercury concentration is below the DNEL_short-value for all the calculated contents of mercury in CFLs.

The calculations are based on an accident with a single compact or straight fluorescent lamp. The calculations thereby indicate that if more than one CFL is broken at the time, the evaporation of mercury is larger and by this the need for ventilation greater.

However, the calculations are based on many uncertainties. Amongst others:

The calculation model assumes that the entire amount of mercury (here 10 % of the total content of Hg in the CFL) evaporates instantly, when the accident happens, as the calculations do not account for the evaporation rate. Mercury does, however, evaporate quickly (7 % is evaporated within a couple of minutes), and therefore the overestimation of using 10 % is not that high.
We have assumed in the calculations that the mercury vapours only are spread within the breathing zone of 2 m³ and not beyond this area, as well as that the vapours are distributed equally in this volume. This assumption can cause an overestimation.
The model assumes that the consumers are exposed to the entire amount of mercury during the entire time of exposure – i.e. the 30 minutes it takes to clean-up (except the amount of mercury that is being removed through ventilation). Therefore, the model does not account for the fact that the real concentration of mercury in the breathing zone will be lower, if the exposure source (the broken CFL) is removed from the room during the exposure time. If you remove the mercury and CFL residues quickly, the calculations are hence overestimated.
The model assumes that mercury vapours are evenly distributed by use of a fan, and therefore ventilation of the room will result in an even ventilation of the mercury vapours in the room. It has not been investigated further, if heavy ventilation is as effective on renewal of the air at the floor where the mercury vapours are concentrated compared to renewal of air higher in the room. However, investigations show that renewal of air from a window also has an effect on the concentration of mercury at the floor.
The DNEL_short-value is calculated by use of a LOAEL-value for an exposure for a couple of hours (not specified further). It is assumed in this report that clean-up in worst case will take half an hour. If clean-up is quick – i.e. 10 minutes, the exposure time will be significantly shorter, meaning that the exposure will be significantly lower than calculated.

The conclusion of short-term exposure to mercury vapour, i.e. during clean-up of a broken CFL is therefore, that when the uncertainties and the assumptions of the calculations are taken into account, there is no health risk, when a CFL breaks at home, if the CFL residues are removed at once and the room is heavy ventilated at once.

Which levels can be found at long-term exposure at accidents with broken bulbs at home?

This situation is a situation where you do not remove all the residues of the broken CFL and therefore can be exposed to mercury vapours on a long-term basis. For this scenario, it has not been possible to carry out a calculation of mercury vapours in the room, as it depends on many factors, such as ventilation, how well clean-up has been performed etc. The evaporation of mercury can in principle continue as long as residues of mercury are left in the room.

This is why the calculated DNEL_long-value is compared to concentrations of mercury vapours, which have been measured by tests with broken CFLs. Measurements have also been performed after clean-up of broken CFLs. These concentrations are described in different tests found in literature.

It is described in the literature that practical tests have proven that left over mercury from a broken CFL can evaporate mercury vapours in a room several weeks/months after the accident – and this even though clean-up was carried out. In some cases it took several weeks before the measured values were below the American long-term concentration without harmful effects at 0.0003 mg Hg/m³(0.3 µg Hg/m³) and thereby also under the calculated DNEL_long-value of 0.0004 mg Hg/m³(0.4 µg Hg/m³).

Extra ventilation after an accident is therefore important – especially in connection with the ordinary cleaning/vacuuming in the home, which can give the result that the dust containing mercury can be whirled up. Ventilation has a considerable effect on lowering the mercury concentrations to non-harmful levels in homes after accidents with a broken CFL.

At long-term exposure to mercury vapours the conclusion is that if all the mercury residues are not removed properly, they may cause a health risk. It is important with thorough clean-up, quickly after the accident happens, as this will remove most of the mercury. Furthermore, it is important to continue the ventilation after an accident, as mercury vapours can be released from non-visible residues of the broken CFL in several weeks/months after the accident.