Sampling Techniques to Ensure Quality Results

David Springer – ACLCA Adelaide June 2013

Tonight

• Measurement Uncertainty

• Poor Sampling Examples

• NEPM Preservations/Holding Times

• NEPM – TRH v’s TPH

• NEPM – F1/F2

• NEPM – BaP TEQ

• NEPM – WA Asbestos

• NEPM – B7

• Transport of Samples – Ice

• Acknowledgement – Thanks to ALS for assisting with some information

Measurement Uncertainty (MU)

• ‘A parameter associated with the result of a measurement that characterises the dispersion of the values that could reasonably be attributed to the measurand’

• (international vocabulary of basic & general terms in metrology)

Simple Definition of MU

• Basically – how sure is the lab of the result?

• It is a range containing the ‘true value’ with a stated level of confidence

– eg: Lead in soil has a MU of 10%. We test a soil & report to you 20 mg/kg.

– We are 95% confident that the concentration of Lead is somewhere between 18mg/kg & 22mg/kg

How to Calculate – various ways

• Precision and Bias

• Last resort – professional judgement

Analyte MU %

Alkalinity 21

Chloride 15

Sulfate 17

Nitrate 14

Nitrite 13

TSS 16

pH 3

Aluminium 18

Cobalt 21

Mercury 16

MU continued

• Depends on matrix and PQL

– eg: OC to 0.001ppm = 80%, to 0.1ppm = 15%

• Small uncertainty – reliable

• Large uncertainty – caution may be appropriate

• Biggest contributor to MU is sampling!

Conundrum

• Lab reports 1 mg/L for Lead

• Guideline value is 1mg/L

• Lab MU is 10% (ie: 0.9 – 1.1 mg/L)

• Is your result over the limit?

Our results are only as good as your Sampling.

• If you don’t get the sampling right, the testing may be a waste of money and your interpretations will be flawed.

How much do we actually take?

• 250g soil jar

• 10g PAH

• 5g BTEX

• 3g Metals

• 40g Asbestos

Foundry Soil

Lead – not too bad

0

500

1000

1500

2000

2500

3000

3500

0 10 20 30 40 50 60

mg

/kg

samples

Pb mg/kg

Series1

Copper - OK

0

100

200

300

400

500

600

700

0 10 20 30 40 50 60

mg

/kg

samples

Cu mg/kg

Series1

Chromium – non homogenous

0

50

100

150

200

250

300

0 10 20 30 40 50 60

mg

/kg

sample

Cr mg/kg

Series1

Pb in top 60mg/kg, Pb in bottom <1mg/kg

Not the best

Water - Field Duplicates

Water Field Duplicates

Duplicates? – Apparently Yes!

Acidification of Metals

0

20

40

60

80

100

120

0 1 3 5 7

pp

b in

so

lutio

n

Elapsed Time (hrs)

The Effect of Time on the Co-precipitation of Metals in a Neutralised Solution Containing 10 ppm Iron and 100ppb of Each Analyte.

Zn

Co

Ni

Cu

Cd

Pb

Filter v no filter - Fe Dropped out of solution

Always take a good QC sample

BaP (PAH) Duplicate - <0.1mg/kg top, 5mg/kg bottom

Hg Hotspots – Lab Duplicate was 4 & 40 mg/kg

NEPM Changes

Analyte Container Holding Time Old NEPM

Holding Time NEW NEPM

Cyanide P, PTFE, G 7 days 14 days

Fluoride P, G 7 days 28 days

Metals P (old) P/G (new) 6 months 6 months

Mercury P (also Cr6) P 28 days 28 days (7 days Hex)

VOC (except below) G 14 days 14 days

Vinyl chloride, styrene, 2-chloroethyl vinyl ester

G 14 days 7 days

TPH v TRH

• 2 new methods – TRH and TRH (silica)

• No longer referred to as TPH

• All labs will now extract/analyse/calculate/report the same way

• Advantages of New Fractions

• More info to make reasonable risk assessments

• Gains of New Method

• Standardisation of banding amongst labs = more consistent data

• Tightening of performance requirements

• The Term TRH is equivalent to the historically reported TPH

Banding

Volatile (vTRH) Semi Volatile (svTRH)

Was C6-C9 Was C10-C36

Now C6-C10 Now >C10-C40

Petrol, light fuel, petroleum based solvents

Diesel, other petroleum fuels, mineral oils and petroleum based solvents

Semi Volatile Fractions

• Interferences

• TRH includes any organic compounds that are soluble in the relevant solvent(s) and elute under a linear GC program method conditions.

• Can include vegetable & animal oils, fats, plasticisers and solvents i.e. Biogenic plus any Petrogenic hydrocarbons

• The use of silica gel to adsorb polar compounds may reduce potential false positives from Biogenic material thus yielding a result more reflective of any petroleum contamination (where applicable).

Fractions

>C10 – C16

>C16 – C34

>C34 – C40

What is Silica Gel Cleanup

• After a sample is extracted in solvent the solvent extract is poured through a column containing silica gel (or can be mixed in the extraction vessel directly). The solvent extract runs down through the column slowly, over the surface of the silica gel particles. The silica gel retains the more polar compounds that were co-extracted from the sample, allowing the less polar petroleum based compounds to flow through ready for GC-FID analysis.

• Note, Silica Gel analysis is only applicable to semi-volatile fractions.

Diesel spiked Woodchips - TRH

C10-C14 : 410ppm

C15-C28: 1500ppm

C29-C36 : 600ppm

Diesel spiked Woodchips - TRH – Silica Gel cleaned

C10-C14 : 380ppm

C15-C28 : 910ppm

C29-C36: 140ppm

The last 2 fractions yield a lower result after the silica gel cleanup, emphasizing the importance of determining TRH – Silica Gel where Biogenic material can lead to false positives.

The term ‘TRH-Silica’ is used as the separation of petrogenic and biogenic is not absolute!

Banding – Old v New NEPM

Old TPH Band Reporting

C10-C14: C>9-C14 : 180ppm

C15-C28: C>14-C28 : 280ppm

C29-C36: C>28-C36 : <PQL

New TRH Band Reporting

>C10-C16: 270ppm

>C16-C34: 190ppm

>C34-C40 : <PQL

The banding is now defined specifically in NEPM B3 with respect to certain n-alkane markers and must be adhered to!

F1, F2

In order to compare to HSLs, F1 and F2 are defined as:-

• F1 = C6-C10 less BTEX (as BTEX HSLs characterised)

• F2 = >C10 – C16 less Naphthalene (as Naphthalene HSL characterised)

From NEPM B1

(CCME)

F1, F2 continued

• F2 = >C10 – C16 less Naphthalene

• Napthalene for this subtraction will be taken from the BTEX analysis (due to BTEX generally being run with TRH). Most labs will do this, though NEPM does not say!

• Otherwise, a separate PAH run (and charge) would possibly occur.

• Always subtraction from BTEX run, even if PAH’s are done – keep consistent.

• Napthalene by P&T (BTEX) v’s Napthalene by GC-MS

• Soils – different extraction (methanol v DCM-Acetone), different instrumentation

• Waters – different preservatives, different instrumentation

WA Asbestos

• WA method = 10x more conservative than the Netherlands to account for dryer Australian soils

• WA = 0.001% asbestos in soil for FA (Fibrous Asbestos) & AF (Asbestos Fines)

• However, 0.01% remains the standard LOR for AS4964 method

• The examination of 500ml may improve the likelihood of identifying asbestos

• Any reporting of <0.01% MUST be NON NATA – Further info TBA

WA Asbestos

• A positive result would normally be considered by DOH to exceed the 0.001% w/w investigation criteria applied to fine asbestos material, especially given that a 0.01% LOD usually applies.

• However, a single such exceedance may not necessarily result in the sample source being deemed contaminated.

• A weight of evidence approach should be used by the consultant, auditor and regulator in assessing the significance of an exceedance, which should take account of the history of the site and frequency and occurrence of other positive and negative results.

Schedule B7 – Free Cyanide

• HIL has been derived for Free Cyanide

• However, the measurement of Free CN is difficult

• So, Schedule B3 recommends WAD CN as a conservative measure of Free CN.

Schedule B7 (Table 1A(1) HIL) - Mirex

• Phased out of Australia in 2007

• Prior to 2007, only used in NT for Giant Termite Control in Mango Trees

• Never Detected before in Australian Food (FASANZ 2001)

• No Background concentration data available in Australia

Schedule B7 - (Table 1A(1) HIL) - Toxaphene

• Mainly available 1940’s to 1980’s

• Still used in some parts of Africa and Asia

• Mainly used in Cotton, Grains, Fruits/Veg, Nuts, Cattle

• NEVER USED in AUSTRALIA !

Benzo (a) Pyrene TEQ

• B(1) includes HIL’s for BaP as Toxic Equivalents (TEQ)

• This is the comparison of 8 carcinogenic PAH’s relative to BaP

• Labs will do this calculation for you

Sample Transport

• Old NEPM – Ice or 4 degree refrigerated transport

• New NEPM – Preferably ice bricks or refrigerated container

• ELIG preferred option has always been –

• Cool samples in either fridge or ice

• Then - transport to lab with ice bricks or double-bagged ice (not free flowing ice)

• Free flowing ice can easily melt, and potentially contaminate your samples.

Potential Cross Contamination

Potential Cross Contamination?

Hydrocarbon Contamination?

Cairns to Sydney does need Bubble Wrap!

Floating in Ice Mud

Ice water penetration - contamination

Esky Ice water

esky water

• Al = 200,000 ppb

• As = 35 ppb

• Cd = 7 ppb

• Co = 930 ppb

• Cu = 3700 ppb

• Fe = 600,000 ppb

• Pb = 120 ppb

• Mn = 55,000 ppb

• Ni = 450 ppb

• U = 24 ppb

esky water + sampling rubbish

• Al = 6000 ppb

• Cu = 15 ppb

• Fe = 7500 ppb

• Pb = 40 ppb

• Mn = 60 ppb

• Zn = 80 ppb

• Ba = 120 ppb

• Sn = 20 ppb

• V = 10 ppb

Metals (total) in melted esky ice

Sydney Sydney Country NSW Sydney West NSW

Aluminium (ppb) 1,300 200,000 5,700 22,000 140

Arsenic (ppb) 2 35 2 20 1

Cadmium (ppb) <0.1 7 0.2 0.9 0.1

Chromium (ppb) 1 120 6 44 5

Cobalt (ppb) 5 930 1 28 <1

Iron (ppb) 1,500 600,000 7,500 24,000 380

Lead (ppb) 1 120 35 49 3

Zinc (ppb) 23 660 77 350 31

Manganese (ppb) 130 55,000 56 540 14

Duct Tape Sealed Bottles

Duct Tape Sealed Bottles

• The melted ice water was analysed for VOC’s:

• Toluene = 520 ppb

• (detection limit is 1 ppb)

Red Texta Stained Ice Water (from label ID’s)

Texta Water – VOC Analysis

• Benzene = 20 ppb

• Toluene = 190 ppb

• Ethyl Benzene = 2 ppb

• Xylenes = 4 ppb

• (Detection limit is 1µg/L = 1ppb)

Esky ice with a sheen?

Esky ice with a sheen - results

• Benzene = 2,100ug/L

• Xylenes – 4,300 ug/L

• Napthalene = 240 ug/L

• C10-C14 TRH = 7,000 ug/L

• C15-C28 TRH = 6,600 ug/L

What’s the Solution?

• Cool samples in ice/fridge – then transport with ice bricks

• Use ice, but double bag it. Don’t leave it free flowing.

What the Lab needs from you

• Clearly written and correct COC’s

• (PB won our 2012 $500 prize for consistently good

COC’s). Over 50% of all COC’s have issues

• Correct sampling – right bottles, metals filtered if needed (and ticked)

• Samples packed to avoid breakages

• Samples packed to avoid cross contamination

• Any doubts – just ask the lab. We don’t mind.

Asbestos or Soil?

Well, I guess we do ask for a separate bag!

No Ice Bricks? At least they made the effort!

Started a trend?

Again?

Ice Brick Fail !

Bacon Brick? – Fail. (the note says `sorry about the bacon J’)

As an Industry – lets get rid of Bad Ice

And only use Good Ice

Photo: Tom Raftery