Skip directly to search Skip directly to A to Z list Skip directly to site content




The Minnesota Department of Health (MDH) was contacted on June 8, 2001 by the Minnesota Pollution Control Agency (MPCA) and Judy Grant, a representative of Xcel Energy Company (Xcel), with concerns about the contamination of a house with mercury. A reportedly small amount of mercury was spilled from a piece of equipment which had been removed from the natural gas line in the basement of the house by Mueller Pipeliners under contract to Xcel. Elemental mercury vaporizes and can be inhaled by individuals in the vicinity of a spill. Exposure to mercury vapors has been associated with permanent and severe central and peripheral nervous system effects.

This health consultation contains a summary of cleanup operations and the health implications of mercury vapor data from this residence. Throughout this document Xcel is cited as the party assuming the obligation to clean up the spill. MDH may have identified some tasks performed by subcontractors, as tasks performed by Xcel.

A Xcel Energy report on the spill and cleanup activities is attached to this health consultation (Attachment #1; Xcel Energy, 2001c). Additionally, a record of mercury vapor concentrations measured in the house is attached (Attachment #2; Xcel Energy, 2001d). These reports contain a summary of daily activity at the site and a compilation of mercury vapor data collected.

Instruments and Methods
Mercury vapor data was collected from a handheld instrument operated by employees of MDH, MPCA, Xcel, and Weston, Inc (Chicago). Table #1 contains a list of the instruments used on the site, operating parameters, and instrument limitations. Information about the Tekran mercury vapor analyzer is also included because it has been used by MDH and MPCA at other residences in Minnesota.

Table #1.

Portable Mercury Vapor Analyzers
Instrument Model Flowrate Sampling Period Detection Limit Sensitivity (resolution) Limitations
(Arizona Instrument, Tempe, Arizona)
431-X 0.75
4 - 13
High detection limit / low sensitivity
(OhioLumex, Cleveland, OH)
RA-915+ 20
(1 ng/m3) sensitivity decreases near 50,000 ng/m3
(Nippon Instrument Corporation, Osaka, Japan)
EMP 1-A 1.5
60 sec
(survey - 1 sec)
(0.001 mg/m3) High detection limit / low sensitivity
(Tekran, Inc., Toronto, Ontario)
2537A 1.5
(0.10 ng/m3) operates from trailer - 60 foot Teflon hose connection

L/min - liters per minute
sec - seconds
mg/m3 - milligrams (1/1000 gram) per cubic meter
ng/m3 - nanograms (1/1,000,000 milligram) per cubic meter

Three different handheld instruments were used to measure mercury vapor concentrations in real-time at the St. Paul house. A Jerome meter was used by BayWest; MPCA, MDH, and Xcel used a Lumex; and Weston used a Nippon. In addition, Applied Environmental Services collected air samples in SKC Hydrar tubes. Air was drawn through the tubes at a rate of 0.5 L/min for about 6.5 hours. These tubes were then analyzed using OSHA ID-140 methods.

Summary of events
The removal of mercury-containing equipment that was attached to natural gas residential service and the subsequent mercury spill occurred on June 5, 2001. The mercury-containing equipment was taken to an Xcel Energy service shop and disassembled. At the time it was disassembled, it contained a couple of tablespoons of mercury (Xcel Energy, 2001c). The device was identified as a low pressure gas regulator or "slam-valve." Previous limited experience with this type of equipment was the basis for Xcel personnel suggesting that similar regulators contained 5 to 6 pounds (lbs) of mercury and were sold as add-on equipment for low pressure protection of individual houses prior to 1950.

Xcel Energy contracted BayWest to clean up the spill in the basement of the residence on June 7, 2001. Cleanup entailed the following:

  • Vacuuming (Minutemen Mercury Vacuum)
  • Mercury cleanup powder (Mercury Absorb, Lab Safety)
  • Wiping areas of concern (Mercon Wipes, Mercon)
  • v
  • Spraying areas of concern (Mercon Vap, Mercon)
  • Ventilating through a basement window with fans.

The maximum mercury vapor concentration measured during this cleanup was reported to be 349,000 nanograms per cubic meter (ng/m3) (Xcel Energy, 2001a). This high concentration was not expected; therefore, questions about possible interference, or cross-sensitivity, from ammonia in a cat litter box were raised. It was confirmed that kitty litter and other odorous material can interfere with correct readings on the Jerome meter.

Following cleanup, BayWest reported mercury vapor concentrations in the basement, with windows open and fans blowing, were "going down to" the Jerome detection limit of 0.001 milligrams per cubic meter (mg/m3; 1,000 ng/m3) (Xcel Energy, 2001a).

Carol Hubbard, of MPCA, and Carl Herbrandson, of MDH, accompanied Judy Grant of Xcel Energy Environmental Services into the house on the afternoon of June 8. Using a Lumex, with a detection limit of 2 ng/m3, mercury levels were measured outside the house, on the main floor, and in the basement. Mercury concentrations in ambient air outside the house fluctuated between less than 30 ng/m3 and more than 150 ng/m3. The high concentrations were recorded when the instrument and the sampling hose were in the sun. The instrument readings appeared to stabilize below 30 ng/m3 when the instrument was held in the shade. Typical mercury concentrations outdoors are 1-3 ng/m3. It is assumed that the elevated readings outdoors were a result of outgassing of mercury from the sampling hose and, perhaps, outgassing of internal parts of the instrument. At a later date it was confirmed that the meter had been recently used in a highly contaminated environment. Apparently, the sampling tube became contaminated, making it difficult to accurately record low mercury vapor concentrations.

We entered the kitchen through the back door and measured mercury concentrations between 1,500 and 3,500 ng/m3 in the kitchen. The mercury concentration increased on the stairs to the basement. Mercury concentrations in the basement ranged from 20,000 to 51,000 ng/m3. According to the manufacturer, the sensitivity of the Lumex flattens out as mercury concentrations in air approach 50,000 ng/m3. Therefore, actual mercury vapor concentrations in the basement on June 8 may have been greater than recorded data.

The Lumex is not thought to be sensitive to ammonia, and readings near the kitty litter box were consistent with what was encountered in other areas of the basement.

The owner of the house was advised, by Xcel, that she would be put up in a hotel since levels of mercury vapor were above levels of health concern (see below). The resident stayed in her house for a couple of nights between June 8 and July 11, when MDH sent her a letter advising her that the house was safe to reoccupy (Attachment #3: MDH, 2001). She spent the other nights at a hotel, or at friends' or relatives' houses.

On Wednesday, June 13, a contractor experienced in cleaning mercury spills in basements (Weston, Chicago, Illinois) was hired by Xcel Energy to clean the residence. Using a Nippon mercury survey meter, Weston recorded a maximum reading of 756,000 ng/m3 during their initial inspection. Moveable items in the basement were bagged and an initial cleanup was performed. Following this cleaning, Weston recorded a maximum mercury vapor reading of 757,000 ng/m3.

Weston continued cleanup operations on June 14 and 15. On June 15 all moveable items were removed from the basement. Cleaning methods included use of mercury cleanup (amalgamating) kits, washing and scraping. Further information about the cleanup can be obtained from the Weston Mercury Cleanup Report (Mueller Pipeliners, 2001). Cleanup continued until all levels of mercury vapor, except those near the floor drain, were below the calibration limit (1,000 ng/m3) on a Nippon mercury vapor survey meter (Xcel Energy, 2001a). On Friday afternoon (June 15th) Weston sealed the floor with an enamel paint. On Friday night the house was vented.

Early on Saturday, June 16, the house was sealed and the furnace turned on to raise the temperature to about 80° F. At about 11:00 AM, Carl Herbrandson of MDH, accompanied the owner and Judy Grant, of Xcel, on what was to be a final inspection prior to collection of air samples for laboratory analyses. Mercury vapor concentrations on the main floor of the house were between 265 and 589 ng/m3 (Lumex mean of 3 - 10 second samples). Mercury concentrations in the basement ranged from 881 ng/m3 to a maximum of 1231 ng/m3. Since these concentrations were above levels of concern, sample collection was postponed and clean up by Weston continued on Wednesday, June 20.

Areas of the floor near the bottom of the basement stairs and a small partition on the north side of the eastern wall, as well as the washtub drain, were identified as potential source areas of mercury vapor. The washtub and partition were removed. Newly applied enamel and some concrete were scraped and/or chiseled from the floor in areas where high mercury readings were recorded. Fast-drying concrete was used to fill cracks and divots in the floor. On June 21, cleaning continued and floor areas associated with elevated concentrations, as well as areas of the walls in the northeast corner of the basement, were then sealed with clear epoxy.

On June 21, Lumex sampling showed mercury vapor levels throughout the house, including the basement, to be below 300 ng/m3. MDH considers a mercury vapor of 300 ng/m3 to be a safe exposure concentration (see below).

On Monday, June 22, air sampling was conducted by Franklin Dickson, Certified Industrial Hygienist, Applied Environmental Services, Inc. (AES). EPA and ATSDR developed protocols (see Attachment #4), used for similarly contaminated houses in Illinois, were used during this sampling. These protocols included closing up the house and heating it to between 75 and 80° F for two hours prior to drawing air samples. Samples were taken in 5 SKC Hydrar tubes (plus one outside sample). Two samples were drawn on each of the first and second floors, and one sample was taken near the laundry area in the basement. A field blank and 2 laboratory blanks were also prepared. The sample tubes were analyzed by Braun Intertech Laboratory. Mercury concentrations throughout the house were below the identified level-of-concern of 300 ng/m3 (see below). Analytical results are listed in Table #2, below (AES, 2001).

Table #2:

SKC Hydrar Tube Data - 6½ hour samples
Sample # Location Mercury Concentration
1. Living Room 210 ng/m3
2. Dining Room 180 ng/m3
3. Master Bedroom 140 ng/m3
4. Second Bedroom 100 ng/m3
5. Basement 270 ng/m3
6. Outdoors - back porch less than 100 ng/m3

On July 3, AES completed their report and faxed it to MDH. Upon review of the AES report and the Xcel summary report, MDH mailed a letter to the resident on July 11, 2001, notifying her that mercury vapor levels in the house were below levels of concern and that it was safe to reoccupy (Attachment #3; MDH, 2001b). On July 13, Carl Herbrandson, MDH, met with Judy Grant and the owner of the house to answer questions and to conduct followup testing with the Lumex (Xcel Energy, 2001a). During this visit, the basement door and windows were closed, but windows on the first and second floor were open. Air flow from outside probably forced the mercury vapor concentrations on the first floor (2 and 16 ng/m3) to levels which may not be indicative of exposure levels in a closed house. But the low levels recorded in the basement (93 to 269 ng/m3) demonstrated a continuing decrease in mercury vapor concentrations in the basement.

Summary of mercury vapor sample data
Differences were noted between readings with the Lumex and the Nippon meters. Direct comparison was made between the 2 mercury vapor meters on June 13 and again on June 15. Side-by-side comparisons are difficult because the Lumex is about 1000 times more sensitive than the Nippon and because the sampling times and the sample flowrates for the 2 instruments are different (see Table #1). Seven of the 11 comparative readings taken with the Nippon on June 15 were at the instrument detection limit (0.001mg/m3). At this level the Nippon would not be expected to be very sensitive to changes in concentration. The remaining readings on the Nippon were between 2 and 5 times the detection limit. Thirty side-by-side samples were taken on the first and second floors on June 13. Another eleven side-by-side samples were taken in the basement (see Attachment #2). The data recorded from the 2 instruments were different. There is no clearly identifiable reason for the differences; however, it is probable that the differences resulted from different sample rates and periods as well as different detection limits (see Table #1). In addition, some data (4) from the Nippon in the basement exceeded 50,000 ng/m3. The Lumex response is known to decrease near this concentration.

MDH and the MCA have confidence that, while the baseline readings may be elevated by contamination of the sampling hose and contamination of the Lumex instrument itself (such as on June 8), the Lumex appears to be accurate and consistent at concentrations above this baseline. The Lumex's accuracy is supported by the close agreement between Lumex data and the data from air samples analyzed in the Braun Intertek laboratory. MDH has not operated Nippon or Jerome meters.

Real-time sampling with handheld instruments was conducted for 2 different purposes: to identify "hotspots" or source areas; and to determine reasonable estimates of exposure to people in the house. Statistical sampling methods were not incorporated when sampling areas of the house. Therefore, while reported mean mercury concentrations in air (Table #3) generally reflect concentrations found in areas of the house, they do not represent true average concentrations. Reported means may overestimate potential human exposures. Maximum recorded concentrations may be close to the actual maximum concentration in an area of the house.

Attachment #2 is a record of data collected at the house by Judy Grant of Xcel Energy. Mercury vapor sampling data were recorded on June 8 (instrument used; Lumex), 13 (Lumex and Nippon), 14 (Lumex and Nippon), 15 (Lumex and Nippon), 16 (Lumex), 18 (Lumex), 20 (Lumex and Nippon), 21 (Lumex), 22 (Lumex and SKC Hydrar tubes), 25 (Lumex), July 3 (Lumex), and July 13 (Lumex).

BayWest did not record mercury vapor concentrations in this house prior to cleanup. Additionally, according to Xcel, BayWest did not submit a written report of their activities to Xcel (Xcel Energy, 2001b). Information about the original cleanup was related to Xcel verbally. Data provided by BayWest are not recorded in Attachment #2. BayWest reported maximum mercury concentrations during their cleanup, on June 7, of about 349,000 ng/m3. While they reported mercury vapor concentrations near the Jerome detection limit (0.001 mg/m3) following their cleanup, mercury concentrations as recorded by Xcel and MDH rose to over 30 times that level the next day, and to over 80,000 ng/m3 within 5 days (see Table #3 or Attachment #2).

Mercury vapor concentrations generally decreased throughout the cleanup. However, there were days, such as June 13, 16, and 20, when concentrations increased (see Table #3 or Attachment #2). On these days it is probable that there was either less ventilation, the temperature was higher, or additional sources of mercury vapor were uncovered or disturbed.

Table #3 is a summary table of available data. Calculated means of Lumex and/or Nippon data are on the left side of the table in bold. Maximum and minimum data are on the right side of the table. Data are categorized as basement, 1st floor and 2nd floor data. Data that were obtained on the stairs to the basement is not included in the table. See Attachment #2 (Xcel Energy, 2001c) for individual sample data.

Mercury Vapor Data (ng/m3) - Means and Ranges
Date Basement 1st Floor 2nd Floor   Basement 1st Floor 2nd Floor
  Mean Mean Mean   Max Min Max Min Max Min
6/8/01 34,876 1,916 432   51,000 23,160 3,525 1,170 674 268
6/13/01 86,114 3,971 1,849   757,000 2,000 16,410 0 2,089 1,687
6/14/01 3,977 216 76   21,000 0 1,000 0 111 46
6/15/01 698 30     2,055 101 41 18    
6/16/01 1,063 428     1,231 881 589 265    
6/18/01 521 213     3,047 60 288 154    
6/20/01 1,506 268     14,770 649 1,000 0    
6/21/01 457 46     1,800 0 337 7    
6/22/01 221 128 109   270 165 171 93 127 92
6/25/01 192 17 7   249 137 37 5 7 6
7/3/01 165 18     212 138 18 18    
7/13/01 139 9     269 93 16 2    

Mercury toxicity health-based exposure values
Health-based exposure values, developed by agencies of state and federal governments, are calculated to be safe exposure concentrations for defined exposure durations. Exposure to chemicals at or below the chronic health-based values for a lifetime, or the acute health-based values for an hour should not result in adverse health effects to the general public including sensitive subpopulations. Protective uncertainty factors have been included in health-based exposure criterion to limit the probability of anyone experiencing an effect, in addition to limiting the severity of any possible effect. However, as the exposure concentration increases above the health-based value, the risk of experiencing an adverse effect increases.

MDH uses health-based reference values from different sources based on availability. Typically, MDH uses health-based numbers in the following preferential order: proposed MDH Health Risk Values (HRVs); US Environmental Protection Agency (EPA) Integrated Risk Information System (IRIS) reference concentrations (RfCs); California Reference Exposure Levels (RELs); EPA Health Effects Assessment Summary Tables (HEAST); and, RfCs and other health-based values, such as Agency for Toxic Substances and Disease Registry (ATSDR) Minimum Risk Levels (MRLs).

Currently there is not a MDH HRV for elemental mercury.

Chronic exposure values: mercury (elemental)
The US EPA IRIS database (2001) specifies a RfC for chronic exposure to mercury vapor of 300 ng/m3. A RfC is an exposure concentration which is not expected to result in adverse health effects in most people, including sensitive subpopulations, exposed over a lifetime. The mercury RfC is based on multiple studies of occupational exposures. Most studies were conducted by studying employees in chlor-alkali plants who were exposed to mercury vapor. The critical effects seen were: hand tremors; increases in memory disturbances; slight subjective and objective evidence of autonomic nervous system dysfunction. The lowest observable adverse effects concentration (LOAEC) in the occupational studies converge at 25 micrograms per cubic meter (µg/m3). (Affected workers had mean whole blood mercury concentrations of 10 - 12 micrograms per liter (µg/L).) Adjusted to a 24 hour, 7 day per week exposure, the LOAECadj = 9.0 µg/m3. An uncertainty factor of 30 was applied to the LOAECadj to reduce the RfC to a level which is assumed to be associated with no adverse effects. The uncertainty factor includes a factor of 10 for human variation in sensitivity, and a factor of 3 for lack of studies on the reproductive and developmental effects of elemental mercury. Therefore, it is presumed that exposure below the RfC will incur no adverse effect.

The calculation of this RfC assumes that there is a threshold level for effects. A threshold for toxicity from mercury vapor exposure is presumed in the standard model used by EPA for non-carcinogens.

The California Office of Environmental Health Hazard Assessment (OEHHA) has adopted a Reference Exposure Level (REL) for chronic inhalation exposure to mercury which is based on the same studies used to develop the IRIS RfC. However, instead of using a cumulative uncertainty factor of 30, which was used by the US EPA, OEHHA has adopted an uncertainty factor of 100. This is based on a factor of 10 for the uncertainty of using a LOAEC exposure instead of a No Observable Adverse Effects Concentration (NOAEC) when calculating the REL, and a factor of 10 for human intraspecies variability. The California REL for mercury (elemental and inorganic) is 90 ng/m3 (CA OEHHA, 2001).

The US Agency for Toxic Substances and Disease Registry (ATSDR) has a health-based minimum risk level (MRL) for mercury of 200 ng/m3 (ATSDR, 1999). This MRL is calculated from the same data that was used to calculate the IRIS RfC. However, the MRL calculation assumes that in an occupational exposure 1/3 of the daily inhaled air each working day is contaminated, whereas the RfC assumes that ½ of the working daily inhalation is contaminated. MDH uses IRIS RfCs for giving exposure advice when there is not a HRV. MDH has some concern that the uncertainty factor of 30 in the EPA RfC, may not be sufficiently protective of a sensitive subpopulation given that the basis of the underlying value is a LOAEL. While the California chronic mercury REL does provide this additional protection, practical application of the mercury REL at contaminated sites may be problematic, since personal exposure to mercury from other sources, including dental amalgams, may be in the range of the REL. MDH therefore recommends that the EPA criterion be used, but that care be taken to ensure that chronic exposures to mercury from all sources do not exceed this level.

Acute exposure values: mercury (elemental)
California OEHHA has developed an acute REL for mercury vapor based on developmental effects in the offspring of exposed rats. Central nervous system effects in pups were noted following exposure of dams to 1.8 mg/m3 for 1 hour/day during gestation. A cumulative uncertainty factor of 1000 is attached to this REL because: it is based on a LOAEL (10X); the primary study was an animal study (10X); and human response to all chemicals is variable (10X). The OEHHA acute REL for mercury vapor is 1,800 ng/m3, with a critical endpoint of reproductive or developmental effects (CA OEHHA, 2001).

Biomarkers of mercury exposure
Biomarkers of exposure may be used to establish a connection between an environmental exposure and an internal level of a toxicant that may cause harm. Biomarkers of mercury exposure are mercury concentrations in whole blood, blood plasma, urine, and hair. The EPA based their mercury RfC on the apparent relationship between whole blood mercury concentrations of 10 µg/L, following chronic exposure to mercury vapor, and adverse health effects. The resident of the house where the spill occurred was examined by a physician, and the physician ordered laboratory analysis of urine, whole blood, and hair. This section contains a discussion of biomarkers of human mercury exposure and the difficulties in interpreting these data.

Exposure to any source of mercury, such as consumption of fish (methylmercury) or inhalation of vapors from dental amalgams, contribute to one's total mercury exposure. Therefore, it can be difficult to relate a concentration in a biological specimen to a specific exposure. The intake of mercury from multiple sources may decrease the utility of biomarkers for exposure to low, yet significant, levels of mercury vapor. This difficulty in associating biomarkers with elevated environmental exposures is typical of human epidemiological studies.

Mercury can be taken into the body in 3 different forms: inhaled as elemental mercury; ingested or dermal absorption of organic mercury; and, ingested or dermal absorption of mercury salts. Exposure to mercury salts is typically related to use of now-banned medicinal creams and teething powders, or occupational exposure. While the largest amount of organic mercury intake comes from the ingestion of fish, there are other potential sources of organic mercury. They include the antiseptics merbromin (mercurochrome) and thimerosal (merthiolate), and dimethyl mercury which is typically found in significant concentrations only in scientific laboratories. Methyl mercury is found in relatively high concentrations in predatory fish; therefore, a diet that includes these fish will result in the intake of potentially significant amounts of methyl mercury. Once in the blood, methyl mercury is transported throughout the body. It can cross through the blood-brain barrier and the placenta, and it accumulates in nervous tissue and the kidneys. Similarly, elemental mercury in the blood can be transported anywhere in the body. As it is slowly converted to inorganic mercury, its mobility becomes restricted.

Whole blood may be a better biomarker of mercury from a diet containing fish than of exposure to mercury vapors. Daily consumption of fish containing 200 µg methyl mercury (very high fish consumption) has been associated with a whole blood concentration of 200 µg/L (WHO, 1990 as cited in ATSDR, 1999), whereas inhalation of 110 µg elemental mercury per day has been associated with a whole blood concentration of 9.8 µg/L (from data supplied in Ngim et al., 1992). Given methylmercury's affinity for organic compounds and disulfide groups, methylmercury in blood is found predominantly in red blood cells (Sandborgh-Englund et al., 1998). It is assumed that inorganic mercury is found predominantly in plasma, and therefore, plasma may be a better indicator of elemental mercury exposure than whole blood (Sandborgh-Englund et al., 1998). Unfortunately, most available biological mercury data are whole blood and urine concentrations. These data may be influenced by an individual's consumption of fish.

Hair can also be used as a biomarker of mercury exposure (Katz & Katz, 1992); however, analysis of mercury in hair is difficult and results can vary significantly between laboratories (Kruse-Jarres, 2000; Seidel et al., 2001). This variability could be the result of differences in pre-analytical preparation of hair samples by different laboratories (Kruse-Jarres, 2000). Additionally, a diet high in fish will lead to high levels of mercury in hair. Therefore, hair may be a better measure of methylmercury exposure than elemental mercury exposure (Wilhelm & Idel, 1996).

It is important to consider that blood mercury is a measure of mobile mercury. Since mercury is sequestered in tissues in the body (e.g., in the kidneys and nervous system) and released slowly (Ozuah, 2000; Sandborgh-Englund et al., 1998; Ekstrand et al., 1998), blood concentrations may not be a measure of a total body burden. Similarly, urine mercury is a measure of mercury excretion and, therefore, cannot be used to determine the total body burden unless measurements are conducted over an extended period of time. An advantage to using hair as a biomarker is that it presumably retains an amount of mercury that is proportional to the concentration in whole blood at a specific time (Toribara, 2001). As a result, even though comparison of hair mercury between individuals may not correlate with exposures, analyses of sequential segments of an individual's hair can be used to develop a historic record of changes in whole blood mercury over time.

Quantitative evaluation of biomarkers of exposure
Human whole blood generally contains less than 10 µg/L (Leavelle, 2001; Ozuah, 2000) mercury. Below 10 µg/L (20 µg/L; Ozuah, 2000) is considered normal urine mercury (Leavelle, 2001). Blood plasma mercury levels in individuals with no known source of inorganic mercury are about 0.22 µg/L (Sandborgh-Englund et al., 1998). While inter-laboratory differences in pre-analytical hair preparation make hair a sub-optimal biomarker, mercury concentrations in hair of less than 1-5 parts per million (ppm; µg/g) have been considered normal (Ozuah, 2000).

Biological measures of exposure from the resident, whole blood mercury and urine, were considered by an occupational physician to be normal (blood mercury below 5 µg/L, urine mercury below 10 µg/L, and hair mercury below 1 µg/g).

Even relatively high exposures to mercury may not be associated with elevated biomarkers. Fifteen minute exposure of 9 volunteers to about 400,000 ng/m3 mercury vapor (220 times the acute REL), led to a median (n=9) maximum whole blood concentration of 1.4 µg/L (at 7.1 hours post exposure), median maximum plasma concentration of 1.2 µg/L (at 8.9 hours post exposure), and median maximum urinary excretion of 0.6 µg/day (at 8.5 days post exposure) (Sandborgh-Englund et al., 1998). In another study chronic exposure to an average of 5,500 ng/m3 elemental mercury for 5.5 years (18 times the chronic RfC), which appears to result in adverse health effects, was associated with a mean whole blood mercury level of 9.8 µg/L (n=98) (Ngim et al., 1992). This is only 2 times a blood mercury concentration that is considered "normal."

Furthermore, differences in individual's diet and the number of dental amalgams can affect mercury biomarkers. Pooling data from available data on the relationship between amalgams and plasma mercury, Ekstrand et al. (1998) calculated that the amalgam contribution to plasma mercury is about 0.02 µg/L/amalgam surface. Therefore, individuals with a typical number of amalgam surfaces (25) may have associated plasma mercury concentrations of 0.50 µg/L. Typical whole blood mercury concentrations in people who do not eat fish are about 2 mg/L (Nordberg et al., 1992).


Exposure to mercury at St. Paul residence
Since being notified of the spill on June 8, MDH has consistently recommended that the resident of this house consult her physician or an occupational physician. Physicians can determine if biomarkers of exposure are elevated, determine if there are specific symptoms of exposure, and they can recommend treatments if appropriate.

It has been reported to MDH that mercury levels in the owner's blood and urine remained within the normal range in all samples taken to-date. Additionally, analysis of hair taken in August showed no elevated levels of mercury. Therefore, there are no positive results from analyses of biomarkers of exposure. As noted above, hair is interesting because it can retain a record of mercury exposure in sequential segments. If hair mercury was elevated, time-correlated samples of segments of individual hairs could have been analyzed to determine a probable exposure scenario.

According to the Environmental Incident Reporting Form, as reported in Attachment #1 (Xcel Energy, 2001a), more than 1 pound of mercury may have spilled from the regulator. If the regulator contained 5-6 pounds of mercury when it was installed, and it contained ‘a few tablespoons' of mercury when it arrived at the service center on June 5, 4-5 pounds of mercury remain unaccounted. This discrepancy has not been addressed in documents summarizing the spill at this house. More mercury being spilled during this event than reported could explain the extremely high mercury vapor concentrations measured in the house during the early cleanup stages. If mercury was spilled or ejected from the regulator prior to the spill on June 5, the residents' exposure to mercury vapor could have been considerably greater than exposure estimates in this document. However, if ambient levels of mercury vapor in the house in the months prior to June 2001 approached levels sampled during the cleanup, MDH would have expected the resident to have elevated biomarkers of exposure.

Relatively high exposures to mercury vapor may have occurred in this house, especially from June 5 through June 13. Exposures during this 9 day period may have been: between 2,000 and 4,000 ng/m3 on the first floor; 400 and 1800 ng/m3 on the second floor; and 35,000 to 86,000 ng/m3 in the basement. It is probable that during this time some individuals, including the resident, were exposed to mercury vapor concentrations above the California acute REL of 1,800 ng/m3. Additionally, exposures above the acute REL may have occurred in the basement for a couple of days after June 13.

As the basement was cleaned, mercury concentrations on the first and second floor, generally, decreased as can be seen in Table #3. This suggests that there was not a significant amount of tracking of mercury out of the basement. Consequently, exposure to mercury vapor on the first and second floor of the house would have been related to the amount of air flow from the basement into the rest of the house, and the exchange of air with the outdoors.

Current daily exposure to mercury in this house is well below levels of concern. Means of measured concentrations on the last sample days listed in Table #3 show ambient mercury vapor concentrations in the basement of 139 ng/m3; first floor, 9 ng/m3; and second floor, 7 ng/m3. If, in a conservative exposure estimate, an individual spends 22 hours a day on the first and second floor, and 2 hours in the basement, about 450 ng of mercury vapor may be inhaled daily. In contrast, published estimates of mercury inhalation from amalgam fillings are 1,000 to 10,000 ng/day (Ekstrand et al., 1998).

Failure to clean up mercury in this house could have led to significant exposure of current and future residents. Using realistic exposure scenarios, the daily exposure to a resident may have been about 200 times the current exposure and 15 times a "safe" exposure calculated using the EPA RfC.

By all measures to date, cleanup of this house appears to be successful.

The resident has expressed some concern that sealing mercury into the basement floors or walls with epoxy will only delay the release of mercury vapor. While MDH shares some similar concerns, it is unlikely that remaining potential sources will become exposed at a single time in the future. Therefore, mercury vapor concentrations should remain below levels of concern for the life of the house. As a precaution, MDH recommends that mercury vapor hotspot sampling be conducted in this house quarterly for at least one year. Furthermore, MDH plans to place this site on a list of houses with cleaned mercury spills for future reference. Future studies of mercury spills in residences may create future interest for mercury vapor monitoring in this house.

Mercury vapor in ambient air in Minnesota residences
In the last 2 years MDH has reviewed mercury vapor data from a number of households as well as in ambient outdoor air. Typical outdoor mercury vapor concentrations are about 2 ng/m3. Air in homes where no known mercury spills have occurred have ranged between 5 and 31 ng/m3 (Tekran; n=5) (MDH, 2001a, in preparation).

MDH is writing a health consultation on another house where 4 thermometers were broken on a carpet (MDH, 2001a, in preparation). The owner cleaned by removing visible mercury and glass from the carpet where the thermometers broke, and vacuuming with a standard vacuum cleaner. When ambient air in the house was sampled 2 weeks later, mercury vapor concentrations were 750 to 800 ng/m3 (Tekran) throughout the house. Two days after the affected carpet was removed, mercury concentrations dropped to about 110 ng/m3 (Tekran). Fourteen months after the incident, in August 2001, followup samplings of mercury vapor were below 10 ng/m3 (Lumex) throughout the house. For additional information, see the health consultation on this house (MDH, 2001a, in preparation).

MDH has also written health consultations for a house where precious metals were recovered from silverware and electronic equipment using cyanide and mercury (MDH, 1998; MDH, 2001b, in preparation). These documents describe decreasing concentrations of mercury vapor in the house over 3 years. Following the initial cleanup in 1998, the maximum concentration measured in the basement was 4,000 ng/m3 and 400 ng/m3 (OSHA Method ID-140) in living areas. The resident was allowed to return to her house because: the resident had no known symptoms of exposure including no elevated biomarkers of exposure; the resident of the house was a widow who wanted to remain in the house; known sources had been removed from the house; mercury vapor concentrations were expected to decrease over time; and, mercury vapor levels in living quarters in the house were near the EPA mercury RfC. When the house was next tested in July 2000, mercury levels had decreased significantly throughout the house. Due to the limited sampling hose length (Tekran - see Table #1), only one area of the basement was sampled at about 4-5 feet above the floor (breathing zone) in 2000. The highest concentration recorded at that location was 166 ng/m3. The highest mercury vapor concentration measured on the first floor at that time was 22.7 ng/m3 (Tekran). When MDH returned to the house in August 2001, mercury vapor concentrations in the breathing zone of the basement ranged from 107 to 147 ng/m3 (Lumex). Higher mercury vapor concentrations were found near the floor. These data are discussed in a separate health consultation (MDH, 2001b, in preparation).

(Health of sensitive sub-populations)

EPA RfC's are developed to be protective of sensitive individuals and children. Children and fetuses have been demonstrated to be sensitive to some mercuric compounds including methylmercury. However, increased sensitivity to elemental mercury has not been studied closely and has not been demonstrated. MDH believes that the uncertainty factors recommended by EPA (30) and OEHHA (100) can be justified. If the maximum concentration in ambient air is kept below 300 ng/m3, exposure to mercury vapor in this house should not affect health.

Because there are many potential sources and routes for exposure to mercury, MDH recommends implementing precautions to minimize these exposures to children and women of child-bearing age.


A private home was contaminated by mercury spilled from a gas regulator on June 5, 2001. Three separate cleanups of the house were undertaken by 2 different companies over a period of 2 weeks. MDH stated that the house was safe for reoccupancy on July 11, 2001.

Exposure to mercury vapor in this house does not now represent a public health hazard. While a spill of mercury from a gas regulator resulted in very high mercury vapor concentrations in this house, the cleanup was successful in lowering levels below those of concern. Further, venting of the house since the cleanup has resulted in additional decreases in mercury vapor concentrations.


MDH is concerned that there may be other houses in Minnesota with mercury-containing gas regulators. Xcel and other natural gas companies in Minnesota have stated that they are not aware of the regular use of mercury-containing gas regulators in Minnesota, nor are they aware of additional houses containing low pressure regulators similar to the one found at this house. Given the high mercury concentrations in this house, and the potential for mercury vapor at these concentrations to cause health effects, MDH will be writing a factsheet on mercury-containing gas regulators and health-related concerns. In addition, MDH will be recommending to Xcel and other natural gas companies operating in Minnesota, that they:

  • remove all mercury-containing regulators from customers' houses over the next few years
  • perform removal with care so as to assure that no mercury is spilled
  • use a sensitive realtime mercury vapor analyzer (e.g. a meter with a detection limit below 100 ng/m3) following removal to assure that residents will not be exposed to elevated levels of mercury vapor
  • confirm that mercury vapor concentrations in indoor ambient air that may have been impacted by a spill from a mercury regulator are below 300 ng/m3
  • notify customers about any exceedances of the EPA IRIS RfC for mercury vapor (300 ng/m3)
  • maintain permanent records on houses from which mercury-containing gas regulators are removed

MDH is planning to make information on mercury-containing gas regulators available on the Web.


MDH recommends that mercury vapor hotspot and ambient air sampling be conducted in this house quarterly for at least one year.

Additionally, MDH recommends that the Public Health Action Plan for this site be adopted and completed.

While MDH has a limited number of recommendations specific to this site, MDH has the following general recommendations.

Recommendations: Characterization of contamination
The following general recommendations are related to the characterization of mercury contamination at indoor spill sites. These recommendations have been reviewed and supported by MPCA staff involved in characterizing mercury contamination in the environment.

  • Sampling for the purpose of determining occupancy of a contaminated site should be conducted using protocols developed by EPA and ATSDR for use in Illinois (Attachment #4). Sampling methods include:
    • controlled ventilation (e.g., no open doors or windows)and no air exchange with outdoor air,
    • indoor ambient temperature greater than 75° F, and
    • sampling in high traffic living areas, breathing zone, and over an extended period of time (typically 6 - 8 hours).
  • Instruments used to characterize contamination should have detection and quantification limits below 100 ng/m3 (e.g. Lumex or Tekran).
  • If there are accompanying concerns about personal exposure and adverse health effects MDH recommends:
    • blood plasma and urine are the preferred biomarkers of exposure.
    • biological sample data should be reported to the analytical detection limit, not an arbitrary reporting limit.
    • hair analyses should be undertaken only if important information may be obtainable from sequential segment analysis.

Recommendations: Site cleanup
Activities and ambient air sampling results described in this document reinforce general MDH recommendations on the cleanup of mercury spills. These recommendations were developed with, and are supported by MPCA staff.

  1. Protect yourself. Before beginning to cleanup a mercury spill:
    • open windows,
    • change into old clothes that can be discarded if mercury spills on them,
    • remove all jewelry,
    • wear gloves and
    • wash thoroughly after cleanup.
  2. DO NOT use a vacuum cleaner or other household cleaning accessories (including brooms) which can cause turbulence or dispersal of small mercury droplets.
  3. Ventilate affected rooms and other rooms sharing common heating and air conditioning systems, during and following cleanup.
  4. Put bulk mercury in a screw-top container, seal it and label it.
  5. Double-bag and label all cleaned-up mercury and contaminated items including tools, gloves and clothing.
  6. Store mercury wastes out of reach of children and away from heat or flames.
  7. Eye-droppers, razor blades, stiff paper, and moist cotton balls can be used to pick up visible mercury contamination.
  8. Sticky tape can remove some mercury residues.
  9. Mercury cleanup kits can be used to collect spilled mercury that is not clearly visible.
  10. Remove and dispose of contaminated carpets and other contaminated permeable materials.
  11. Dispose of unused mercury-containing thermometers, bulk mercury, bagged items, and other items containing mercury (including spill cleanup aids, items and tools) at an appropriate hazardous waste facility.
  12. For help with a mercury spill call the MPCA through the Minnesota Duty Officer at:
    Twin Cities metro area (651) 649-5451
    Statewide (800) 422-0798


Carl Herbrandson, Ph. D.
Site Assessment and Consultation Unit
Environmental Surveillance and Assessment Section
Minnesota Department of Health


Applied Environmental Sciences, Inc. (2001) Environmental Health Survey Report: Mercury clearance testing. Dickson, F., AES, Minneapolis, MN. 7/3/01.

Agency for Toxic Substances and Disease Registry (1999) Toxicological Profile for Mercury. Research Triangle Institute, ATSDR, U.S. Department of Health and Human Services, Public Health Service, Atlanta, GA. March 1999.

California Office of Environmental Health Hazard Assessment (2001). Reference Exposure Levels. Office of Environmental Health Hazard Assessment, CA OEHHA, Sacramento, CA. (1/22/01).

Ekstrand, J., Bjorkman, L., Edlund, C., Sandborgh-Englund, G. (1998) Toxicological aspects on the release and systemic uptake of mercury from dental amalgam. Eur J Oral Sci 106:2 Pt 2 p678-86.

Environmental Protection Agency, U. S. (2001) Integrated Risk Information System. Online, Office of Health and Environmental Assessment, Environmental Criteria Office, EPA IRIS, Cincinnati, OH

Katz, S. A., Katz, R. B. (1992) Use of hair analysis for evaluating mercury intoxication of the human body: a review. J Appl Toxicol 12:2 p79-84.

Kruse-Jarres, J. D. (2000) Limited usefulness of essential trace element analyses in hair. Am Clin Lab 19:5 p8-10.

Leavelle, D. E. (2001) Mayo Medical Laboratories Interpretive Handbook. In. Mayo Press, Rochester, MN, pg 456-458.

Minnesota Department of Health (1998). Chemically Contaminated South Minneapolis Residence. Minneapolis, Hennepin County, Minnesota. Herbrandson, C., MDH, ATSDR Concurrance: June 11, 1999.

Minnesota Department of Health (2001) Herbrandson, C. Letter to: Resident, Concerning: Mercury levels 'safe' to reoccupy house. July 11, 2001.

Minnesota Department of Health (2001a, in preparation). Mercury in ambient air in a Minnesota house: Contamination from broken thermometers. Marine, Washington County, Minnesota. Herbrandson, C., MDH, Prepared in collaboration with ATSDR:

Minnesota Department of Health (2001b, in preparation). Chemically Contaminated South Minneapolis Residence: Followup. Minneapolis, Hennepin County, Minnesota. Herbrandson, C., MDH,

Mueller Pipeliners (2001) Mercury clean up: XXXXXXXXXXXXXXXX, St.Paul, MN, 13-21 June 2001. prepared by: Roy F. Weston Inc., Mueller Pipeliners, New Berlin, WI. July 2001.

Ngim, C. H., Foo, S. C., Boey, K. W., Jeyaratnam, J. (1992) Chronic neurobehavioural effects of elemental mercury in dentists. Br J Ind Med 49:11 p782-90.

Nordberg, G., Brune, D., Gerhardsson, L., Grandjean, P., Vesterberg, O., Wester, P. O. (1992) The ICOH and IUPAC international programme for establishing reference values of metals. Sci Total Environ 120:1-2 p17-21.

Ozuah, P. O. (2000) Mercury poisoning. Curr Probl Pediatr 30:3 p91-9.

Sandborgh-Englund, G., Elinder, C. G., Johanson, G., Lind, B., Skare, I., Ekstrand, J. (1998) The absorption, blood levels, and excretion of mercury after a single dose of mercury vapor in humans. Toxicol Appl Pharmacol 150:1 p146-53.

Seidel, S., Kreutzer, R., Smith, D., McNeel, S., Gilliss, D. (2001) Assessment of commercial laboratories performing hair mineral analysis. Jama 285:1 p67-72.

Toribara, T. Y. (2001) Analysis of single hair by XRF discloses mercury intake. Hum Exp Toxicol 20:4 p185-8.

World Health Organization (1990) Methyl Mercury. Vol 101: (as cited in ATSDR, Toxicological Profile for Mercury), WHO, International programme on chemical safety, Geneva, Switzerland.

Wilhelm, M., Idel, H. (1996) Hair analysis in environmental medicine. Zentralbl Hyg Umweltmed 198:6 p485-501.

Xcel Energy Company (2001a) Activity and Information Summary June 5 - July 13, 2001: XXXXXXXXXXXXXXXXXXX, St. Paul MN. Grant, J., Xcel Energy, Environmental Services, Minneapolis, MN. July 25, 2001.

Xcel Energy Company (2001b) Gerlach, M. Email: Herbrandson, C., Minnesota Department of Health. Concerning: Availability of reports. August 9, 2001.

Xcel Energy Company (2001c) Grant, J. Phone call to: Herbrandson, C., Minnesota Department of Health. Concerning: Mercury contaminated House, St. Paul. June 8, 2001.

Xcel Energy Company (2001d) Mercury vapor data - June 5 - July 13, 2001: XXXXXXXXXXXXXXXXXXXX, St. Paul MN. Grant, J., Xcel Energy, Environmental Services, Minneapolis, MN. July 25, 2001.


This Mercury in a St. Paul Residence Health Consultation was prepared by the Minnesota Department of Health under a cooperative agreement with the Agency for Toxic Substances and Disease Registry (ATSDR). It is in accordance with approved methodology and procedures existing at the time the health consultation was begun.

Alan W. Yarbrough
Technical Project Officer, SPS, SSAB, DHAC, ATSDR

The Division of Health Assessment and Consultation, ATSDR, has reviewed this public health consultation and concurs with the findings.

Sven E. Rodenbeck
Chief, Superfund Site Assessment Branch, DHAC, ATSDR

Next Section     Table of Contents The U.S. Government's Official Web PortalDepartment of Health and Human Services
Agency for Toxic Substances and Disease Registry, 4770 Buford Hwy NE, Atlanta, GA 30341
Contact CDC: 800-232-4636 / TTY: 888-232-6348

A-Z Index

  1. A
  2. B
  3. C
  4. D
  5. E
  6. F
  7. G
  8. H
  9. I
  10. J
  11. K
  12. L
  13. M
  14. N
  15. O
  16. P
  17. Q
  18. R
  19. S
  20. T
  21. U
  22. V
  23. W
  24. X
  25. Y
  26. Z
  27. #