BRANCH OF QUALITY ASSURANCE MEMORANDUM 90.03
SUBJECT: ADP - Storage of water-quality, quality-assurance data in NWIS

Increasing numbers of water-quality quality-assurance (QA) samples (blanks, blinds, spikes) are being submitted for analysis from field activities. Because the National Water Information System (NWIS) was not designed to accommodate these types of samples, most of the QA data are not being stored in the data base. Those Districts, or projects, that have stored QA data in the data base were forced to develop their own systems or protocols to keep track of that data. The Branch of Quality Assurance would like to consolidate these data into a national data base so that systematic trends or errors in sampling techniques, preservation methods, and laboratory analyses might be detected. Until the Districts store these data in a consistent manner, the effort to consolidate a national data base is impractical.
The Branch of Quality Assurance (BQA) has developed a protocol, programmed modifications, and arranged for new parameter codes to enable the 90.2 release of the NWIS-I software to store and retrieve QA data. These procedures are designed for storage of field-derived QA data only, and will not accommodate internal, laboratory, quality- control samples. The QA samples are identified by new medium codes and sample types. The QA samples may be related to specific environmental samples by using the same station number, date and time. New parameter codes and new remark codes are used to identify further the type of data stored with each QA sample.
An alternate data base for quality-assurance data is recommended to prevent inadvertent retrievals of the QA data. The alternate data base must be named QAFILE to be properly identified by the new programing procedures. When QAFILE exists, the data-entry programs will automatically separate QA analyses (as identified by medium code) and enter them into QAFILE. If QAFILE does not exist, all analyses, including QA samples, are entered into the regularly assigned data base. Standard retrieval and processing programs in the NWIS system can be used with QA data and the QA data base. Environmental samples and quality assurance samples may be retrieved from their respective data bases and tabled together, using the new programs designed to table data from multiple data bases that are included in the 90.2 NWIS-I release.
The efforts to establish and maintain the QA data base is minimal when done in conjunction with normal water quality data base management activities. One additional management consideration will be the creation and maintenance of lists of artificial site ids used to identify certain trip blanks, reference materials, or spike solutions. Although it is not mandatory to establish a QA data base, the benefits of being able to store, retrieve, and manipulate these data with the environmental data in NWIS ought to justify its existence. In addition, the ability of BQA to pool localized QA data to a national scale ought to help every district and project; further justifying the establishment of individual district QA data bases. The more historical QA data that can be entered into the data base, the more viable that data base becomes. Although new monetary and/or manpower resources to load all historic data into district data bases are not available, BQA will consider aiding specific district requests for inputting special or "significant" historical data.
Chapter 6 of the NWIS water-quality user documentation describes the system for storing quality assurance data. For further information, contact Alan Burns (AWBURNS; FTS 776-3612) or Terry Schertz (TLSCHERTZ; FTS 776-1491) at the Branch of Quality Assurance in Golden, Colorado.
William J. Shampine
Chief
Branch of Quality Assurance
Enclosures:
(1) Definitions of quality assurance data
(2) Identification of quality assurance data in NWIS-I
(3) Lists of new codes
Distribution: A, B, S, FO, PO
Definitions of Quality-Assurance Data
_____________________________________
Blank samples are taken to ensure that environmental samples have not been contaminated by the data-collection process. Any measured value/signal in a blank sample for an analyte (a specific component measured in a chemical analysis) that was absent in the blank solution is believed to be due to contamination. There are many types of blank samples possible, each designed to segregate a different part of the complete data- collection process including sampling, filtering, preserving, storing, transporting, and analyzing:
Blank solution - Solution that is free of the analyte(s) of interest. Such a solution would be used to develop specific types of blank samples as described below.
Shelf (or hold) blank - a blank solution that is put in the same type of bottle used for an environmental sample and stored adjacent to an environmental sample in a storage area.
Refrigerator blank - a blank solution that is put in the same type of bottle used for an environmental sample and stored adjacent to an environmental sample in a refrigerated storage area.
Trip blank - a blank solution that is put in the same type of bottle used for an environmental sample and kept with the set of sample bottles both before and after sample collection.
Sampler blank - a blank solution that is poured or pumped through the same field sampler used for the collection of an environmental sample.
Filter blank - a blank solution that is filtered in the same manner and through the same filter apparatus used for an environmental sample.
Splitter blank - a blank solution that is mixed and separated using a field splitter in the same manner and through the same apparatus used for an environmental sample.
Preservation blank - a blank solution that is treated with the same preservatives used for an environmental sample.
Field blank - a blank solution that is subjected to all aspects of sample collection, field processing, preservation, transportation, and laboratory handling as an environmental sample.
Blind sample - A sample submitted for analysis whose composition is known to the submitter but unknown to the analyst. A double blind sample is one of known composition that is submitted to the analyst in such a manner that neither its composition nor its identification as a check sample is known to the latter. A blind sample is one way to test the proficiency of a measurement process. Blind samples can be used to monitor the performance of an analytical system, check the analytical results of more than one laboratory, more than one analytical method, or the consistency of the same laboratory and method. Every blind sample analyzed should have an associated reference to the source and the possible dilution. Blind samples may be prepared from a reference material, as defined below.
Reference material - A material or substance one or more properties of which are sufficiently well established to be used for the assessment of a measurement method or for assigning values to materials.
Replicate (Duplicate) samples - A group of samples, collected in a manner such that the samples are thought to be essentially identical in composition. Replicate is the general case for which duplicate is the special case consisting of two samples. There are many types of replicate samples possible, each of which may yield slightly different results in a dynamic hydrologic setting, such as a flowing stream. The types of replicate samples are:
Concurrent samples - samples collected by two or more people collecting samples simultaneously, or by one person alternating subsamples between two or more collection bottles.
Sequential sample - a type of replicate sample in which the samples are collected one after the other, typically over a short period of time.
Split sample - a type of replicate sample in which a sample is split into subsamples contemporaneous in time and space.
Spike sample - A sample to which known concentrations of specific analytes have been added in such a manner as to minimize the change in the matrix of the original sample. Every spiked sample analyzed should have an associated reference to the spike solution and the volume added.
Spike solution - A solution with one or more well established analyte concentrations that is added in known quantities to an environmental sample to form a spike sample.
Identification of Quality-Assurance Data in NWIS
________________________________________________
A system for identifying quality-assurance samples and maintaining the relationship with corresponding environmental samples has been established. The MEDIUM CODE, DATE, TIME and SAMPLE TYPE are used in the following manner to identify clearly all QA data:
Blank -
(1) Use actual STATION NUMBER, DATE and TIME for blanks associated with a specific site and environmental sample. Use artificial STATION NUMBER, DATE and TIME for blanks that are associated with many sites and/or environmental samples. If more than one blank is analyzed, increment the TIME by 1 minute for each blank.
(2) Use a MEDIUM CODE of Q (new code).
(3) Use a SAMPLE TYPE of 2 (new code).
(4) Use new parameter codes to designate which type of blank solution, the source of the blank solution, and the type of blank sample.
(5) If more than one blank is collected, use parameters 82073 and 82074 to enter the same STARTING TIME and ENDING TIME for each blank.
Blind -
(1) Use actual STATION NUMBER, DATE and TIME for blinds associated with a specific site and environmental sample. Use artificial STATION NUMBER, DATE and TIME for blinds associated with many sites and/or environmental samples. If more than one blind is analyzed, increment the TIME by 1 minute for each blind.
(2) Use a MEDIUM CODE from Q-Z (new codes).
(3) Use a SAMPLE TYPE of 4 (new code).
(4) Use new parameter codes to designate the source of the reference material and the source code number.
(5) If more than one blind is collected, use parameters 82073 and 82074 to enter the same STARTING TIME and ENDING TIME for each blind.
Reference Material -
(1) Assign an artificial STATION NUMBER.
(2) Use the DATE and TIME that material was created or received.
(3) Select a MEDIUM CODE from Q-Z (new codes).
(4) Use a SAMPLE TYPE of 6 (new code).
(5) Use new parameter codes to designate the source of the reference material and the source code number.
Replicate -
(1) Use the existing STATION NUMBER.
(2) Use actual DATE and TIME for the 1st sample and increment TIME by 1 minute (or actual times if greater than 1 minute) for each additional sample.
(3) Select a MEDIUM CODE from Q-Z (new codes).*
(4) Use a SAMPLE TYPE of 7 for every sample, including the first one.
(5) Use a new parameter code for replicates to designate which of the methods was used to create the replicates.
(6) Use the same STARTING TIME and ENDING TIME (parameters 82073 and 82074) for all samples.
* If the replicate samples will be stored in the regular data base, use normal MEDIUM CODES (i.e., 6 or 9) and a sample type of 7.
Spike -
(1) Use actual STATION NUMBER, DATE and TIME for spikes associated with a specific site and environmental sample. Use artificial STATION NUMBER, DATE and TIME for spikes that are associated with many sites and/or environmental samples. If more than one spike is analyzed, increment the TIME by 1 minute for each spike.
(2) Select a MEDIUM CODE from Q-Z (new codes).
(3) Use a SAMPLE TYPE of 1.
(4) Use new parameter codes to designate the source code number of the spike solution, the spike type, and the volume of the spike. Use an existing code (32000) to designate the sample volume.
(5) If more than one spike is collected, use parameters 82073 and 82074 to enter the same STARTING TIME and ENDING TIME for each spike.
Spike Solution -
(1) Assign an artificial STATION NUMBER.
(2) Use the DATE and TIME that source was created or received.
(3) Use a MEDIUM CODE of Q (new code).
(4) Use a SAMPLE TYPE of 8 (new code).
(5)Use new parameter codes to designate the source of the spike solution and the source code number of the spike solution.
A DISTRICT PROCESSING STATUS code of Z (local use only) should be assigned to each QA sample to prevent the data from being transferred to the national data base and subsequently to the EPA's STORET data base.
New Codes
_________
Medium Codes
Q Quality-Assurance Sample -- Artificial
R Quality-Assurance Sample -- Surface Water
S Quality-Assurance Sample -- Ground Water
T Quality-Assurance Sample -- Wet Deposition
U Quality-Assurance Sample -- Bulk Deposition
V Quality-Assurance Sample -- Suspended Sediment
W Quality-Assurance Sample -- Animal Tissue
X Quality-Assurance Sample -- Plant Tissue
Y Quality-Assurance Sample -- Interstitial Water
Z Quality-Assurance Sample -- Bottom Material
Sample Types
1 *Spike
2 Blank
3 *Reference
4 Blind
5 *Duplicate
6 Reference Material
7 *Replicate
8 Spike Solution
B Other Quality Assurance
* Existing sample types that apply to quality assurance data
Remark Codes
A Average value
V Variance
S Most probable value
Parameter Codes
99100 Blank, Type of solution (fixed value)
10.00 Distilled/deionized Water
20.00 Standard Reference Water Sample
30.00 Matched Matrix
40.00 Organic-free water
200.00 Other
99101 Blank, Source of solution (fixed value)
10.00 National Water Quality Lab
20.00 U.S. Environmental Protection Agency
30.00 Standard Reference Water Sample (USGS)
35.00 Mix of Standard Reference Water Samples
40.00 NIST (formerly NBS)
50.00 Canadian Inland Waters
60.00 District lab
70.00 Natural sample
80.00 Ocala Lab
99.99 Unknown
100.00 Chemical Supplier
200.00 Other
99102 Blank, Type of sample (fixed value)
10.00 Shelf (hold)
20.00 Refrigerator
30.00 Trip
40.00 Sampler
50.00 Splitter
60.00 Filter
70.00 Preservation
100.00 Field
200.00 Other
99103 Reference Material, Source (fixed value)
10.00 National Water Quality Lab
20.00 U.S. Environmental Protection Agency
30.00 Standard Reference Water Sample (USGS)
35.00 Mix of Standard Reference Water Samples
40.00 NIST (formerly NBS)
50.00 Canadian Inland Waters
60.00 District lab
70.00 Natural sample
80.00 Ocala Lab
99.99 Unknown
100.00 Chemical Supplier
200.00 Other
99104 Reference Material or Spike Source, Code Number
99105 Replicate, Type (fixed value)
10.00 Concurrent
20.00 Sequential
30.00 Split
40.00 Split-concurrent
50.00 Split-sequential
200.00 Other
99106 Spike, Type (fixed value)
10.00 Field
20.00 Laboratory
30.00 Surrogate
40.00 Internal Standards
200.00 Other
99107 Spike, Source (fixed value)
10.00 National Water Quality Lab
20.00 U.S. Environmental Protection Agency
30.00 Standard Reference Water Sample (USGS)
35.00 Mix of Standard Reference Water Samples
40.00 NIST (formerly NBS)
50.00 Canadian Inland Waters
60.00 District lab
70.00 Natural sample
80.00 Ocala Lab
99.99 Unknown
100.00 Chemical Supplier
200.00 Other
99108 Spike Volume