Guidance for Interpretation of NWQL Schedule 2020 and 2021 VOC Data 


To: "A  - Division Chief and Staff",
        "B  - Branch Chiefs and Offices",
        "S  - Special Distribution for Research",
        "FO - State, District, Subdistrict and other Field Offices",
        "PO - Project Offices",
        wqspecs@usgs.gov, owq@usgs.gov,
        " , NAWQA Leadership Team, Reston, VA ",
        " , NAWQA National Synthesis Chiefs, Reston, VA ",
        " , NAWQA 91 Study-Unit Project Chiefs, Reston, VA ",
        " , NAWQA 94 Study-Unit Project Chiefs, Reston, VA ",
        " , NAWQA 97 Study-Unit Project Chiefs, Reston, VA "
from: "Janice R Ward, Acting Chief, OWQ, Reston, VA "
cc: " , WRD Archive File, Reston, VA "
Subject: OFFICE OF WATER QUALITY TECHNICAL MEMORANDUM 98.04

Date: Wed, 15 Apr 1998 09:39:57 -0400
Sender: "Nana L Frye, Secretary (OA), Reston, VA "


In Reply Refer To: 
Mail Stop 412                                       April 14, 1998

OFFICE OF WATER QUALITY TECHNICAL MEMORANDUM 98.04

Subject:  Guidance for Interpretation of NWQL Schedule 2020 and 2021 VOC 
          Data 

BACKGROUND

On July 29, 1997, the National Water Quality Laboratory (NWQL) announced 
in NWQL Technical Memorandum 97.11 that a new method for the 
determination of volatile organic compounds (VOCs) had been approved as a 
U.S. Geological Survey method. This method determines 86 VOCs at lower 
reporting levels than were previously available, and is available as 
Schedule 2020 which includes a search for unknown compounds and as 
Schedule 2021 which does not search for unknown compounds. Details and 
pricing of the method are available via the Water Resources Division 
Intranet from the NWQL Home Page.

For pricing and the list of constituents and reporting levels, click on 
"Spin" and request "2020" under a "Schedule" search. For a brief 
description of Schedule 2020, page down to "Technical Information", find 
"Schedules: Information and Frequently Asked Questions" and click on 
"Schedule 2020."

PURPOSE OF MEMORANDUM

This VOC analytical method was available as Lab Codes 9090 and 9091 
before being approved as NWQL Schedules 2020 and 2021. Based on a careful 
analysis of data and procedures, data produced on or after October 1, 
1996, are considered approved USGS data and are suitable to be included 
in annual data reports. Data produced before October 1, 1996, using 
custom Lab Codes 9090 and 9091 remain provisional and may not be included 
in data reports. This memorandum presents guidance on how data from the 
custom Lab Codes may differ from that of the approved Schedule, and how 
those data may be used. The National Water Quality Assessment Program 
(NAWQA) was the principle user of Lab Codes 9090 and 9091.

GUIDANCE AND INTERPRETATION

Provisional data, such as data from custom Lab Codes 9090 and 9091, may 
be included in interpretive reports by referencing or describing the 
analytical method in the interpretive report. Connor and others (1998) 
describe the analytical method and its performance in Open-File Report 
97-829. In addition, the attachment to this memorandum contains two 
paragraphs that can be used in interpretive reports to describe the 
custom method used before October 1, 1996. Data from custom methods 
cannot be included in data reports.

In both custom Lab Codes 9090 and 9091, and approved Schedules 2020 and 
2021, concentrations of analytes were determined only when each analyte 
had been qualitatively determined to be present. Then concentrations were 
determined using valid calibration curves. When concentrations are 
reported for individual analytes from Lab Codes 9090, 9091, or Schedules 
2020 or 2021, these concentrations may be used together for all 
hydrologic interpretations. The difference between data produced using 
Lab Code 9090 or 9091, and Schedule 2020 or 2021, is how low-level 
detections were censored in relation to laboratory blanks.

Between April 1996 and October 1, 1996, for Lab Codes 9090 and 9091, some 
environmental sample detections of common laboratory contaminants were 
censored, regardless of daily blank values.   These included: benzene, 
ethylbenzene, toluene, xylenes, 1,4-dichlorobenzene, methylene chloride, 
acetone, and carbon disulfide.

After October 1, 1996, a more complex and rigorous blank evaluation 
procedure is used for data produced using Schedules 2020 and 2021. This 
procedure results in substantially less censoring of common laboratory 
contaminants. It compares additional quality-control blank sample results 
to environmental sample results to determine whether the environmental 
sample results can be considered valid. These comparisons are described 
in Connor and others, 1998. This more rigorous procedure should show an 
increase in the reporting of common laboratory contaminants after October 
1, 1996, for Schedules 2020 and 2021, compared to results reported before 
October 1, 1996, for custom Lab Codes 9090 and 9091.

In terms of interpretation of environmental results, if benzene,
ethylbenzene, toluene, xylenes, 1,4-dichlorobenzene, methylene chloride, 
acetone, or carbon disulfide were determined to be present at a site 
after October 1, 1996, and the same site yielded a nondetect (less-than) 
value before that time, this difference could be an artifact of the 
laboratory data evaluation procedures. Because of this, counting 
detections both before and after October 1, 1996, can be misleading.

In order to count and compare numbers of detections over the period 
before and after October 1, 1996, all comparisons of the numbers of 
detections must exclude benzene, ethylbenzene, toluene, xylenes, 
1,4-dichlorobenzene, methylene chloride, acetone, and carbon disulfide. 
If those compounds are excluded, then comparisons of numbers of 
detections before and after October 1, 1996 should be valid. However, 
data users are cautioned not to over-interpret presence versus absence of 
compounds at low concentrations.

Another caution to interpreting data at low concentrations is that as 
concentrations decrease near the detection level, the relative standard 
deviation becomes quite large. Therefore, data users are cautioned not to 
over-interpret small concentration differences near reporting limits. To 
interpret data at low concentrations, examine patterns of data rather 
than individual quantitative measurements.

REFERENCE

Connor, Brooke, Rose, Donna L., Noriega, Mary C. Murtagh, Lucinda K., 
  Abney, Sonja R. 1998. Methods of Analysis by the U.S. Geological Survey 
  National Water Quality Laboratory-Determination of 86 Volatile Organic 
  Compounds in water by Gas Chromatography/ Mass Spectrometry, Including 
  Detections Less Than Reporting Limits. U.S. Geological Survey, Open-
  File Report 97-829, 78 p.



				Janice R. Ward /s/
				Acting Chief, Office of Water Quality

This Memorandum refers to National Water Quality Laboratory Technical 
Memorandum 97.11.


Distribution:	A, B, S, FO, PO
		District Water-Quality Specialists
		Regional Water-Quality Specialists
		OWQ Staff
		NAWQA_nlt, NAWQA_nst, NAWQA_91pc, NAWQA_94pc, NAWQA_97pc

Key Words: VOC, 2020, 2021, 9090, 9091, data interpretation

---------------------------


ATTACHMENT

The following is a brief description of custom volatile organic compound 
(VOC) Lab Code 9090 and 9091 analytical method for inclusion in the 
experimental section of U.S. Geological Survey interpretive reports.

Water samples were analyzed by purge and trap capillary-column gas 
chromatography/mass spectrometry at the U.S. Geological Survey's National 
Water Quality Laboratory. Analyses were performed according to Connor and 
others (1998). The analytical method provides reliable evidence of an 
analyte's presence and is suitable for reporting concentrations at 
sub-microgram-per-liter levels for 86 volatile organic compounds (VOCs). 
The method also allows for reporting estimated concentrations for other 
tentatively-identified VOCs. The analytical method began to be used in 
April 1996, as a custom method and all data produced after October 1, 
1996 are approved USGS data. Data produced prior to this date are 
provisional. However, the principle difference between data produced 
before and after October 1, 1996 is how low-level detections of VOCs 
affected by laboratory contaminants were evaluated or censored in 
relation to internal laboratory blanks.

Between April 1996 and October 1, 1996, a different evaluation and 
reporting strategy was used for those VOCs affected by laboratory 
contaminants. Environmental-sample results found to contain laboratory 
contaminants were censored (that is, reported as less than the detection 
limit) regardless of the magnitude of daily laboratory-blank values. 
Additional laboratory blank samples were incorporated into the analytical 
sequence after October 1, 1996, to help distinguish between true 
low-level environmental concentrations and low-level environmental 
concentrations caused by laboratory background contamination. Laboratory 
contaminants that were censored were benzene, ethylbenzene, toluene, 
xylenes, 1,4-dichlorobenzene, methylene chloride, acetone, and carbon 
disulfide. The automatic censoring used between April 1996 and October 1, 
1996, resulted in an underestimate of the occurrence of the censored 
compounds in environmental samples.