Introduction to Model 1130/35 Automated Ambient Air Speciated Mercury Monitor

1
Introduction to Model 1130/35Automated Ambient
Air Speciated Mercury Monitor

• F.H. Schaedlich

Rev 1.11 November 3, 2003
2
Acknowledgements

• Matt Landis US EPA 1
• Robert K. Stevens Florida DEP 2a
• Tom Atkeson Florida DEP 2b
• Eric Prestbo Frontier Geosciences 3
• Steve Lindberg Oak Ridge Natl Labs 4
• Gerald Keeler University of Michigan 5

• 1 Research Triangle Park, NC 27711
• 2a At US EPA, Research Triangle Park, NC 27711
• 2b Tallahassee, FL
• 414 Pontius Ave. N., Suite B, Seattle, WA 98109
• PO Box 2008, Oak Ridge, TN 37831-6038
• 5 School of Public Health, 109 Observatory
  Street, MI 48109.

3
Total Gaseous Mercury Monitoring
Part 1
4
Why measure parts per Quadrillion of Mercury ?

• Mercury is one of the most potent neurotoxins
  known
• Bio-accumulates up the food chain by factors of
  up to 10,000,000 times
• Sub-ppt levels in air can accumulate to toxic ppm
  levels in fish
• Levels are increasing worldwide
• Long life in the atmosphere means that mercury
  emissions are of global concern

5
Total Gaseous Mercury Monitoring

• Originally implemented manually
• Gold coated silica field cartridges
• Analysis using dual stage thermal desorption with
  CVAFS (atomic fluorescence) detection
• Gold preconcentration with AA (atomic absorption)
  detection had been shown to be subject to serious
  artifacts when attempting to measure ambient
  levels

6
Model 2537A Mercury Analyzer

• Automated implementation of gold/AF manual method
• Instrument protected by US and international
  patents
• Has largely supplanted manual cartridge methods
• Provides continuous total gaseous (TGM) readings
  with update rate as low as 2.5 minutes
• Detection limit lt 0.1 ng/m3 (5 min. samples)
• Automatic recalibration with internal Hg0
  permeation source
• Capable of unattended operation for extended
  periods

7
Model 2537A Analyzer
8
Principles of Operation

• Mercury in sample gas is preconcentrated onto
  (patd) pure gold cartridge
• Adsorbed mercury is thermally desorbed
• Detected by atomic fluorescence detector
• Two cartridges are used to alternately sample and
  desorb
• No gaps in data stream

9
Flow Diagram of Model 2537A Mercury Analyzer
(Patd)
10
Pure Gold Cartridge (Patd)

• Pure gold only is used as adsorbent
• No quartz wool or silica
• No memory effect
• Extremely durable design
• Lasts for years of continuous use
• Cartridge design is protected by separate US and
  international patents

11
Pure Gold Cartridge
12
Atomic Fluorescence

• Much more sensitive than atomic adsorption
• MDL lt 0.1 pg absolute
• Not subject to interferences
• AA requires some sort of compensation/correction
  scheme
• Interfering compounds (e.g. SO2, O3, organics)
  often present in concentrations thousands of
  times higher than Hg
• Inherently linear
• Detector linear over gt5 orders of magnitude

13
Tekran AF Detector Features

• Most sensitive, most stable and lowest noise AF
  detector available
• Some reasons why
• Temperature controlled lamp block
• Optical feedback loop for constant lamp intensity
• Optical path purged with argon for stability
• lt 10 ml/min required for optical purge
• Magnetic shield on photomultiplier tube

14
Tekran AF Detector
15
Internal Permeation Source

• Allows automated, unattended calibrations
• Ensures that analyzer results are always accurate
• Allows standard additions to sample gas
• Ensures that analyzer is working properly even
  when confronted with difficult matrices
• Long life permeation tube
• Low emission rate Approx 30 µg/yr
• Theoretical life thousands of years!
• Special auto-flush feature ensures no carryover
  after activation

16
Perm Source Flow Diagram
17
Integrates into Monitoring Networks

• Designed to look and operate like other
  continuous ambient air monitors
• Rack mountable
• Easy to use interface
• Two analog chart outputs
• Serial output provides hi-resolution results and
• Instrument performance variables
• Full record of each calibration
• Record of instrument setup parameters

18
Operational Characteristics

• Lamp lasts in excess of one year of continuous
  operation
• Inexpensive, multi-sourced
• Unlimited shelf life
• No special isotopes required
• No zero drift !
• Each reading is an integrated peak value
• Detector drift does not affect results

19
Reporting Basis

• Reports data readings based on sample volumes
  corrected to 0C, 760 mm
• All Hg concentrations in the literature are
  reported on this volume basis
• No reading-by-reading temperature and pressure
  corrections required to normalize data
• Reporting at other conditions (e.g. 20 C) is
  possible simply by changing internal menu setting

20
True Ambient Measurement

• Tekran Model 2537A is the only analyzer available
  that is capable of accurately and precisely
  measuring true ambient mercury values
• Can measure mercury in actual gas matrices
• Virtually any analyzer (even uncompensated AA
  units) can give correct answers when measuring
  mercury in zero air!

21
Other Applications

• With appropriate accessories and front ends the
  Tekran Model 2537A can perform
• Process gas monitoring
• Continuous emissions monitoring (CEM)
• Mercury flux measurement
• Both chamber gradient methods!
• Standard additions
• Speciation ionic (reactive gaseous) mercury
• Speciation particulate bound mercury

22
Sample Data Chlor-Alkali Plant

• Plot of wind direction vs. ambient mercury values
• Two months of continuous monitoring
• Plant is located 18 miles distant
• Fish consumption advisories in effect
• Fish levels were gt 1.5 ppm even though air
  readings were only slightly elevated
• Global background1.5 ng/m3

23
Depletions of Mercury in the Arctic
The 2537A is capable of extremely precise
measurement of even sub-ambient readings. In
cases where two instruments are running, they
track each other very closely.
24
Validation of Depletion Events
This line shows the recovery of periodic
automated standard additions of elemental mercury
to the sample matrix. The recoveries are 100,
yielding absolute confidence that these unusual
values represented a genuine, hitherto unknown
phenomenon.
25
Reactive Gaseous Mercury Monitoring
Part 2
26
Tekrans Objectives

• Development of an automated method to measure and
  differentiate different forms of mercury in
  ambient air
• Reactive Gaseous Mercury (RGM)
• Fine Fraction Particulate Mercury (HgP)
• Elemental mercury (Hg0)

27
Why Ambient Air Mercury Speciation ?

• Different forms of gaseous Hg have vastly
  different behaviors and environmental impacts
• Forms can interconvert in the atmosphere and in
  various reservoirs
• Classes of mercury in ambient air
• Elemental mercury Hg0, GEM
• Reactive mercury HgII, RGM, Hg2
• Particulate mercury HgP, TPM

28
Elemental Mercury Hg0

• Typically 90 of atmospheric Hg loadings
• Relatively inert. Long residence time
• Hg0 sources impact large areas
• Sources chlor-alkali plants, gold and Hg mining,
  thermal power plants

29
Reactive Mercury Hg(II) or RGM

• Consists of ionic, water soluble forms, primarily
  HgCl2
• Usually only a few percent of total gaseous
  mercury present in the atmosphere
• Short range deposits relatively close to the
  source of emission
• Primary sources Coal burning power plants, waste
  incinerators, chlor-alkali plants

30
Difficulties Measuring RGM

• Method must be 1-2 orders of magnitude more
  sensitive than total mercury methods
• Reject much larger elemental component
• Must exclude particulate bound mercury, however,
  conventional particulate filters cannot pass RGM
• Apparatus must pass RGM to the collector
  quantitatively. (RGM is extremely sticky)

31
Model 1130 - Principles of Operation

• Patented by Tekran. (US 6,475,802)
• Quartz, KCl coated annular denuder is thermally
  desorbed and regenerated
• Sampling Phase
• Absorbs RGM while passing all elemental Hg
• Model 2537A reads Hg0 during this phase
• Desorption/Analysis Phase
• Zero air used as carrier
• RGM released as elemental mercury

32
RGM Sampling Phase

• Denuder is slightly warm
• Inlet slightly warm
• Denuder captures all RGM while passing Hg0
• Total sample flow is 10 lpm
• Analyzer measures Hg0 during this time

33
Analysis Phase

• Denuder is heated to 500C
• Inlet heated to clean up-stream glassware. (75
  C)
• Denuder releases captured RGM as Hg0
• Excess zero air flow and heat cleans upstream
  components
• Instrument measures accumulated RGM as Hg0

34
Quartz Denuder Assembly
35
Model 1130 Speciation System

• Denuder Module (top)
• Located out of doors
• Contains denuder element within high performance
  oven
• Impactor inlet removes coarse particles (gt 2.5
  um)
• Climate controlled enclosure

36
Model 1130 Speciation System

• Pump Module (bottom)
• Located beside 2537A
• Provides extra flow required during sampling
• Generates the zero air required during desorption
• Also provides all zero air required by Model
  2537A!
• Heated line (not shown)
• 25 ft. length connects the two modules

37
Instrument Characteristics

• Fully automated, unattended operation
• Denuder exchanged every 1 or 2 weeks
• Operates in any environment
• Polar (Greenland, USA, Canada, Antarctica)
• Sub-tropical (Florida Everglades)
• Desert (Nevada)
• Marine (ocean cruises)
• High altitude (Mauna Loa)

38
Data Characteristics

• High time resolution Hg0 data provided during
  sampling phase
• Typically every 2½ or 5 min
• Data is averaged to yield one Hg0 point per
  sampling period
• Single RGM integrated value every sample period
• Sample period 0.5 3 hours
• Analysis Period 25 60 min
• Data values are for the exact same time period

39
Method Validation

• Method was developed over a two year period at
  Tekran before being announced
• Tested using both HgCl2 and HgI2 as surrogates
  for RGM
• Two prototype units bought by State of Florida
• Units evaluated by Frontier Geosciences in
  Seattle under contract with Florida DEP
• Subsequent field testing by Keeler et al,
  University of Michigan Lindberg et al., ORNL

40
Lab RGM Apparatus - (1997)
41
Preliminary Results - Toronto
42
Preliminary Results - Toronto
Two curves have been shifted horizontally for
visibility
43
Lab Testing Frontier Geosciences
44
Seattle RGM Data
Municipal Incinerator Plume
Courtesy Frontier Geosciences
45
Mercury Levels - Rural Michigan
Courtesy UMAQL
46
Particulate Mercury Monitoring
Part 3
47
Why Measure HgP ?

• Particulate bound mercury (TPM / HgP) is created
  directly by some industrial processes and also
  formed downwind of sources by the combination of
  mercury with existing particulates
• HgP is removed even more rapidly via wet and dry
  deposition than is RGM

48
Particulate Mercury TPM

• Consists of various compounds of mercury bound
  onto particles
• Size range Generally thought to be lt 3 um
• Usually only a few percent of total mercury
  present in the atmosphere
• Short range deposits relatively close to the
  source of emission

49
Difficulties Measuring HgP

• Method must be 1-2 orders of magnitude more
  sensitive than total mercury methods
• Must reject much larger elemental component
• Method must also exclude reactive gaseous
  mercury, however, normal particulate filters have
  problems in that they retain some RGM
• Virtually all conventional measurement methods
  have recently been found to have serious
  analytical artifacts due to RGM (Landis, et al,
  EST, 2002)
• Artifact is non-quantitative and cannot be
  subtracted or corrected

50
Model 1135 - Principles of Operation

• Sampling Phase
• Sample passes through impactor to eliminate
  coarse particles (gt 2.5 um)
• Coated denuder captures RGM and eliminates RGM
  artifact
• Quartz, Regenerable Particulate Filter (RPF)
  captures fine fraction (lt 2.5 um) particulates
• Desorption/Analysis Phase
• Zero air used as carrier
• Downstream pyrolyser is activated first
• RPF is then heated to desorb captured particulates

51
RPF

• Quartz wool captures bulk of particulate matter
• Quartz filter membrane provides sharp cut-off
  (lt0.1 um)
• Built in pyrolyzer section converts desorbed
  species to Hg0
• Carbon based particulate matter on the filter is
  oxidized to CO2 during desorption cycle
• Tail provides cooling

52
Model 1135 Particulate Mercury Monitor

• Unit stacks on top of Model 1130
• Allows simultaneous deter-mination of Hg0, RGM
  and HgP
• Detection limit lt 2 pg/m3

53
Model 1130 Model 1135 Flow Path
54
Ambient Air Speciation
55
Sample Installations
Figure 1 - Temporary Model 1130/35 Installation
(AES, Quebec Canada)
Figure 2 - Permanent Arctic Model 1130
Installation (NOAA, Barrow Alaska)
Figure 3 - Model 1130-P (prototype) Mounted on
roof of trailer (US EPA, NERL)
Figure 4 - Temporary Model 1130-P Installation
(US EPA Florida DEP)
Figure 5 - Model 1130-P sampling in Antarctica
(GKSS AWI, Germany)
56
Conclusion

• Measurement of mercury species at ambient levels
  is now practical
• Equipment has shown its utility and reliability
  in a wide variety of sampling situations around
  the world
• Some additional commitment in resources is
  required to implement a successful speciated
  mercury monitoring program

57
For More Information

• Contact fhs_at_tekran.com
• Visit www.tekran.com