Process monitoring – feedback for water treatment may improve the efficiency of the treatment process – use fewer chemicals to achieve require effluent emission limits
Measurement of Se, As for compliance with Steam Electric Generating Effluent Guidelines
Monitoring of treatment of wastewater by biological based treatment systems
Measurement of Corrosion Products (e.g. Fe, Ni, Cr and Mn) to improve operating efficiency
Measurement of Elements in Nuclear Power Plant Applications (e.g. Pb, Cu, Fe)
Reduce laboratory analysis costs
3. Why
Measure
Metals in
RealTime in
Water
Treatment
Processes
3
• Process monitoring – feedback for water
treatment may improve the efficiency of the
treatment process – use fewer chemicals to
achieve require effluent emission limits
• Measurement of Se, As for compliance with
Steam Electric Generating Effluent
Guidelines
• Monitoring of treatment of wastewater
by biological based treatment systems
• Measurement of Corrosion Products (e.g. Fe,
Ni, Cr and Mn) to improve operating
efficiency
• Measurement of Elements in Nuclear Power
Plant Applications (e.g. Pb, Cu, Fe)
• Reduce laboratory analysis costs
4. XRF BasedWater
Analyzer Basics
• Xact 920 Utilizes ED-XRF as analytical technique
• Water is spray dried and sampled onto filter tape
• The resulting filter tape deposit is analyzed by XRF
• Builds on two technologies developed and
commercialized by Cooper Environmental
• Ambient Air XRF Analyzer (over 100 in field)
• Quantitative AerosolGenerator (Developed
for Calibration of PM CEMS for the Power
Industry)
• Instrument can measure up to 65 elements
simultaneously
4
5. XRFTheory
• X-ray source is electrically powered NOT a
radionuclide source
• Incoming X-rays eject an inner shell electron
• Electrons from higher shells fill the vacancy
• This process releases energy in the form of
fluorescingX-rays
• Energy is characteristic of each element
• Intensity or brightness is related to the mass of
each element
5
6. Strengths of
XRF
• XRF utilizes inner shell electron transitions so
the response is not dependent on what the
element is chemically bound to
• Can measure a wide range of elements
simultaneously
• XRF is non-destructive – so samples can be
reanalyzed later
• XRF is very stable – calibrations can last for
years
• XRF response is linear over the a wide
concentration range (over 5 orders of
magnitude) – this means no additional
standards required depending on concentration
range
6
8. General Operation Schematic - Xact 920
LBM
Pre-filter Sterilizer
Pump
Drying
Chamber
Internal
Standard
Carbon Trap
Flow Meter
Pump
Exhaust
Filters
Modified
Xact 625
Software Data
Processing/
reporting
Sample
Source
Sample pre-treatment
and transport
Sample
Analysis
Quantitative liquid
blending and sample
pre-concentration
DAP
Data acquisition
and processing
Xact 295
8
Sample is
aerosolized and
dried
Sampled onto filter tape and analyzed by XRF
LBM: Liquid Blending Module – this is where we combine the liquid sample flow with
the a liquid flow containing an internal standard
10. XRF Sampling and Analysis
X-Ray Tube
Filter Tape
Dried Aerosol
Deposit
Dried Aerosol from Drying Chamber
Analysis Area
Filter Tape
11. General Quantification - Xact 920
11
Internal
standard
Effluent
Quantitatively
blended
effluent and
internal
standard
Liquid Blending Module
XRF metals analysis
Compressed air
Blend
sample/
internal
Std.
12. Xact 920 – Quality Assurance
• XRF Portion Calibrated withTraceable to NIST
Thin Film Standards
• XRF spectrometer QA with every sample
• XRF StabilityCheck with Every sample
(Pd Rod)
• Daily upscale check of XRF
• Multi element upscale is inserted
once/day
• Includes energy calibration of detector
• Stability – XRF calibration frequency
~once/year – sometimes years between
calibrations
12
14. Xact 920 Detection Limits
14
15 30 60 120
S 3.8 3.7 1.3 0.47
Cl 2.0 2.0 0.72 0.25
K 1.4 1.4 0.49 0.17
Ca 0.36 0.35 0.12 0.044
Ti 0.19 0.19 0.066 0.023
V 0.14 0.14 0.050 0.018
Cr 0.14 0.14 0.048 0.017
Mn 0.17 0.17 0.059 0.021
Fe 0.20 0.20 0.070 0.025
Co 0.16 0.16 0.056 0.020
Ni 0.11 0.11 0.039 0.014
Cu 0.092 0.091 0.032 0.011
Zn 0.077 0.077 0.027 0.010
As 0.073 0.073 0.026 0.0091
Se 0.094 0.093 0.033 0.012
Br 0.12 0.12 0.042 0.015
Cd 2.9 2.9 1.0 0.36
Pb 0.15 0.15 0.052 0.018
Element
Sample time (minutes)
Detection Limit in ppb
68% Confidence, less then 50 ppm ofTDS
• 30 minute DL’s are less than
1ppb for most elements
• Detection Limit is a function of…
• Sampling time – more time
better detection limits
• Total Dissolved Solids (TDS)
• HighTDS can limit the
amount of preconcentration
that takes place – too much
material on tape can limit air
flow
15. Minimum Detection Limits (ppb)
(Standard Configuration)
0.0
1.0
2.0
3.0
4.0
0 5000 10000 15000 20000
DetectionLimit(ppb)
Dissolved Solid Conc. (ppm)
Xact 920 Detection Limits
60 minute Sample & AnalysisTime
C
r
C
u
A
s
15
*One sigma interference free detection limits
*Detection limits based off standardXact setup. Detection limits can be optimized based off elements of interest
Detection Limit Increases as
TDS increases
Limited by amount of flow
that can pass through the
tape at high aerosol
concentrations
16. Minimum Detection Limits (ppb)
(Standard Configuration)
0.00
0.05
0.10
0.15
0.20
0.25
0.30
0 100 200 300 400 500 600 700 800 900 1,000
DetectionLimit(ppb)
Dissolved Solid Conc. (ppm)
Xact 920 Detection Limits
60 minute Sample & AnalysisTime
Cr
Cu
As
Pb
Se
16
*One sigma interference free detection limits
*Detection limits based off standardXact setup. Detection limits can be optimized based off elements of interest
Detection Limit Remains constant when sampling is not limited byTDS
26. Field Demonstration
• Field test done at one of the world’s largest metal melting
facilities
• Water from several plant processes and run-off from
snowmelt and water sprayed to reduce fugitive emissions
• MetalsTreatment included adding lime slurry to raise PH
(~10.5) and a flocculant to precipitate metals
• Sampling done downstream of clarifier
• Instrument operated at facility for over four months
26
28. Xact 920-ICP-MS Comparison
Multi-Element Standard
28
• After instrument installed in the field the instrument was tested using
a multi element standard
• Gravimetric Results – how the Xact compared with the known
concentration of the solution
• ICP-MS – How the Xact compared with analysis off line by Inductively
Couple Plasma Mass Spectroscopy
29. Xact 920-ICP-MS Comparison
“As Found” Grab Sample
29
• Average percent difference less than 20% for all elements exceptTl
• Good agreement between ICP-MS and the Xact
• In generally good stability in the analysis results from the Xact –
the relative percent standard deviation is less than 20% for all
elements except copper
• …. Indicates values lower than the LOQ for the Xact
LOQ= Limit of Quantitation
30. Measurement
of Spiked
Samples
30
• Effluent samples were spiked with several
elements of interest at two concentration
levels (30, and 300 ppb)
• The resulting solutions were measured by
the Xact 920 and by ICP-MS
• The percent difference between the Xact
and ICP-MS was determined
• Percent recovery for the Xact was calculated
31. Xact 920-ICP-MS Comparison
Spiked Grab Sample
31
• Results within 20% of ICP-MS for all elements
• Percent relative standard deviation less than 10% for all
elements indicating good precision in the measurement
32. Xact 920 Spike Recovery Results
32
• Percent recovery calculated for all elements – a value of 100% would
represent a perfect result
• EPA 200.8 Criteria for spiked recovery is 100% +/- 30% - All of the elements
met this criteria
34. Coal Mine
Grab Sample
Experimental
Setup
34
ANALYZED GRAB
SAMPLE FOR TOTAL SE
ANALYZED SPIKED
GRAB SAMPLE FOR
TOTAL SE
SUBMIT RESULTS TO
CLIENT
CLIENT SENDS US
REGULAR LABORATORY
TOTAL SE ANALYSIS
RESULTS
COMPARE RESULTS
35. Coal Mine Grab Sample – Xact
920 vs Lab Analysis
35
Conc. (ppb) σ (ppb) Conc. (ppb) σ (ppb) *
% σ*
Grab sample (sampled 11/7/2016) 25.6 4.3 31 2 -17 15
*
CES assumed an uncertainty of 5% for TRL analysis
Sample Info
Avg. Xact 920 Analysis Lab. Analysis Percent Error
• Xact 920 and Laboratory result agree within 20%
Se Spiked Conc.* Measured Se Conc. σ Se Conc. Spike Recovery Stdev Spike recovery RPD
ng/g ng/g ng/g % % %
022317_48 Grab sample as received 30.3 3.1
022317_49 Grab sample as received 21.9 2.3
022317_50 Grab sample as received 24.6 2.5
Average Grab sample as received 25.6 4.3**
021717_40 Grab+50 ng/g Se spike 51.3 87.2 9.0 120.3
021717_41 Grab+50 ng/g Se spike 51.3 80.7 8.3 107.5
022017_44 Grab+200 ng/g Se spike 197.6 201.0 6.5 88.9
022017_45 Grab+200 ng/g Se spike 197.6 208.9 6.8 93.0
Note:
* Total Se (solution consists of 89.03% Se6+ and 10.97% Se4+)
** Standard deviation of three measured replicates
9.0 11.2
2.9 4.4
Sample Information Xact 920
Run ID Sample Description
N/A N/A N/A N/A
• Spiked recovery results well within EPA recommended range of 70% to
130%
36. Conclusions
36
XRF Based Metals
Water Analyzer is
able to measure a
wide range of metals
at concentrations
down to less than 1
ppb
Provides results that
are comparable to
those obtained by
laboratory analysis
Is robust enough to
operating in
demanding field
conditions
Notas do Editor
Zero drift = abs(measured value – initial measurement)/Full scale range, largest drift over period of time equals zero drift percentage
Range drift = abs(measured value – initial measurement)/Full scale range, largest drift over period of time equals zero drift percentage