PROGRAMS AND PLANS:  Implementation of the Protocol for Collecting and Processing Surface-Water Samples for Low-Level Inorganic Analyses

In Reply Refer To:                                  July 15, 1993
Mail Stop 412      


Subject:  PROGRAMS AND PLANS:  Implementation of the Protocol 
          for Collecting and Processing Surface-Water Samples 
          for Low-Level Inorganic Analyses


For 2 years, the Office of Water Quality (OWQ) has been developing 
a new field protocol for the collection and processing of samples 
for low-level inorganic analyses.  The purpose of this memorandum 
is to announce the implementation date for the protocol and to 
address issues of concern to Water Resources Division (WRD) 
personnel.  Several issues are covered: (1) reasons for the 
protocol; (2) date of implementation; (3) what is meant by 
"implementation"; (4) programs and projects affected, including 
logistical considerations and equipment needs; (5) training needs; 
and (6) answers to some specific questions that frequently have 
been asked.

                   REASONS FOR THE PROTOCOL

The protocol was developed and is being implemented to improve and 
assure the quality of data produced by WRD in the operational 
program.  It addresses a recognized need for improvement in 
methods to enable WRD to meet its mission of describing the 
quantity and quality of water resources.  Through employment of 
the protocol, WRD will produce more accurate, quality-assured, 
inorganic data.

More specifically, the necessity for the protocol is detailed in 
a series of OWQ Technical Memorandums, beginning with 91.10, 
describing problems with the quality of existing trace-element 
data for filtered samples.  In that memorandum, two key points 
were made.  First, research projects in several geographic areas 
demonstrated that filtered trace-element concentrations for many 
elements are well below 1 microgram per liter (ug/L) (down to
a few nanograms per liter for certain elements).  Comparing some 
of the research data to National Stream-Quality Accounting Network 
(NASQAN) results implied probable contamination in the U.S. 
Geological Survey data.  Second, studies on equipment blanks and 
actual field samples demonstrated an unacceptable level of 
contamination from random and systematic sources--mostly from 
sample collection and field processing--for at least eight trace 
elements.  Based on these findings, the OWQ determined that (1) 
changes were needed in field methods, field equipment, field 
supplies, and associated quality assurance practices; (2) field 
quality-control (QC) sampling needed to be instituted; and (3) 
increased QC was needed on blank water and other supplies.  The 
resultant protocol is the summation of all of the needed changes 
and refinements.

                     IMPLEMENTATION DATE

October 1, 1993, is the date the WRD will implement the new 
protocol.  By that time, the OWQ will have (1) developed and 
tested the protocol, (2) identified appropriate equipment and 
supplies needed to implement the protocol, and (3) established a
new laboratory method to analyze large numbers of filtered samples 
for low levels of trace elements (using a new method for 
inductively coupled plasma/mass spectrometry (ICP/MS) together 
with existing methods).  Protocol documents currently in 
circulation are Drafts.  The OWQ will issue a Technical Memorandum 
to formally present the protocol as Division policy on October 1, 

                    IMPLEMENTATION DEFINED

Implementation means that from October 1, 1993, onward, WRD will 
use the protocol for all programs and most projects.  

For new programs and projects developed after this date, the 
protocol will be the standard operating procedure (SOP) for the 
production of inorganic data.  The protocol will be the SOP for 
collecting and processing samples for major ions and nutrients, as 
well as trace elements.  Thus, for new projects, the protocol 
should be used whenever inorganic samples are collected, even if 
trace-element samples are not collected (see "Commonly Asked 
Questions" for a more detailed discussion).  A few projects where 
only major ions are collected or where work is being done in a 
highly contaminated environment may not require use of the 
protocol.  These projects must generate suitable and adequate QC 
data to ensure that the environmental data being produced are not 

For existing programs and projects, the protocol should be applied
as the SOP as soon as possible.  This implementation may take 
up to a year to implement in all WRD programs and projects.  
NAWQA will begin using the protocol--as modified for ground 
water--in a pilot manner this summer.  NASQAN and Benchmark will 
begin using the protocol as SOP at the beginning of fiscal year 
(FY) 1995 (October 1, 1994).  This timing will coincide with 
implementation of a redesigned NASQAN program.  However, Districts 
that wish to do so are encouraged to begin using the protocol for 
NASQAN and Benchmark stations as soon as possible.  The OWQ will 
work with such Districts to help implement the protocol, but 
cannot pay for new sampling equipment and additional supplies.

                       PROGRAMS AFFECTED

All WRD operational programs and projects are affected by 
implementation of the protocol.  As described above, after 
October 1, 1993, all new projects should use the protocol as 
the standard SOP.  For existing programs or projects, full 
implementation will depend on several logistical considerations. 
Most aspects of the protocol (such as the equipment cleaning) 
are relatively easy and inexpensive to implement.  Those portions 
should be initiated as soon as possible.  In addition, generation 
of QC field blanks is both necessary and comparatively simple to 
implement.  The field QC data will verify if contamination 
problems exist that need to be addressed.

The protocol lists sampling equipment and supplies that OWQ has 
tested and established as capable of meeting protocol requirements 
(see attachment 1).  Appropriate samplers are available now from 
the Waterways Experiment Station and supplies will be available 
from the Ocala Quality Water Service Unit by October 1, 1993.  

Because the equipment tested to date does not cover all sampling 
environments, full implementation of the protocol is impossible 
under all conditions (for example, under ice sampling).  However, 
to improve data quality, OWQ suggests implementation of 
as much of the protocol as practical as quickly as possible.  


As part of the protocol testing, one District per Region has 
received training and has agreed to provide future training to 
other personnel in use of the protocol.  Other Districts should 
arrange training through the Regional Water-Quality Specialist. 

Field training has shown that the protocol (1) is easy to 
follow, (2) adds little time at a site when two-person sampling 
crews are used and equipment cleaning is done in the office, and 
(3) can be effectively implemented with about 2 days of training 
followed by some practice.  The basis for success is development 
of an awareness of the potential for contamination and the desire 
to carefully employ the procedures.  Field personnel who are 
implementing the protocol are doing so enthusiastically because of 
measurable improvement in the quality of data.

                    COMMONLY ASKED QUESTIONS

Question 1:  This protocol is for surface-water sampling--
    shouldn't it include ground-water sampling?

Answer:  Ground-water sampling is being addressed through several 
    efforts concurrent to development of this surface-water 
    protocol.  A draft document titled "Protocols and procedures 
    for collection of ground-water-quality data" by Wilde, Lapham, 
    and Koterba contains information parallel to that found in the 
    surface-water protocol.  In addition, the OWQ is developing a 
    National Field Methods Manual that will contain succinct 
    protocols for surface-water and ground-water sampling.  Thus, 
    expanding this protocol to include ground-water sampling would 
    be duplicative.

Question 2:  Trace elements are reported at ug/L levels, but
    nutrients and major ions are reported at milligram per liter 
    (mg/L) levels.  Why does this protocol include nutrients and 
    major ions?

Answer:  There are two reasons why nutrients and major ions were 
    included in the OWQ testing and subsequently in the protocol. 

    First, several nutrient schedules have 0.01 or 0.001 mg/L
    reporting limits.  These are actually ug/L reporting levels,
    and many ambient conditions include nutrient concentrations
    in this range.  The cleaning, QC, and other items included
    in the protocol are necessary to produce good-quality nutrient
    data at these levels.  

    Second, given that the protocol is necessary to ensure good- 
    quality data for nutrients and especially for trace elements, 
    it would be counterproductive for most programs and projects 
    to use a separate protocol--based on different equipment and 
    supplies--for collecting samples just for major ion analyses.  

Question 3:  Many samples are collected for unfiltered analyses; 
    the protocol is for filtered samples.  How does the protocol
    apply to unfiltered samples?

Answer:  The focus for protocol development was on filtered 
    samples because these have the more rigorous requirements
    to reduce contamination levels.  However, experiments clearly 
    show that the sample collection step is a major source of
    contamination for a number of trace elements (see OWQ
    Technical Memorandum 92.05).  Therefore, on samples having 
    low to moderate concentrations of suspended sediment, use of 
    the protocol on unfiltered samples will be necessary to avoid 
    the production of contaminated trace-element data.  For 
    samples having high concentrations of suspended sediment, the 
    concentration of trace elements extracted from the sediment 
    will swamp contamination added in the sample collection step. 
    Unfortunately, the actual level of suspended sediment at which
    contamination is overwhelmed by extracted concentrations of 
    trace elements will vary by element, sampling site, time of 
    year, and other factors. Because the actual level will never 
    be known at the time of sampling, commencing October 1, 1993,
    the protocol will be the Division SOP for all unfiltered as 
    well as filtered samples collected for inorganic analyses.

    In a related matter, in 1992, the National Water Quality 
    Laboratory (NWQL) instituted an in-bottle digestion method 
    for unfiltered samples that has greatly reduced the potential 
    for contamination while increasing the precision of results.

Question 4:  Are two-person sampling crews always necessary?

Answer:  Two people are needed to do the "clean hands/dirty 
    hands" technique.  Districts that use two-person crews have 
    found that they (1) produce better quality data, (2) save 
    time, and (3) are worth the cost.  Aside from improperly 
    cleaned equipment, use of one person will probably be the 
    greatest potential source of contamination in sample 
    collection and processing.  Based on field experience of WRD, 
    university, and other agency scientists, single-person 
    sampling from bridges, cableways, and metal boats will produce 
    contaminated samples.

    In addition to the contamination issue, two-person crews 
    promote safety.

Question 5:  Do I have to spend $2,000 on a D-77?

Answer:  For some sampling circumstances, there are other, less 
    expensive alternatives to D-77 samplers.  For wading 
    situations, a DH-81 with a plastic bottle is acceptable and 
    inexpensive.  The metal rod can be covered with an inexpensive
    piece of electrical-type shrink tubing.  For low-velocity 
    situations (< 1.5 ft/s), a weighted-bottle sampler can be 
    used.  HIF has a prototype for a noncontaminating weighted-
    bottle sampler.  For sampling from a bridge or a boat, a 
    frame-type, bag sampler can be fabricated for about $150.  The 
    OWQ will obtain drawings for the frame samplers and provide 
    them to all Districts prior to October 1, 1993.

    Use of other sampler types is not certified by the OWQ because 
    (1) the designs allow potential contact of water with metal 
    surfaces, and (2) tests conducted and reported by OWQ show 
    consistent contamination at unacceptable levels.  OWQ testing
    has shown the samplers included in the protocol to be 
    acceptable for low-level inorganic sampling.  If you want to 
    use other samplers, you are responsible for generating 
    adequate tests and QC data to substantiate that your 
    objectives can be met with each other sampler.

Question 6:  Dedicating a vehicle specifically for water quality 
    may be financially infeasible under all conditions in every 
    District.  Is this necessary?  Are there alternatives?

Answer:  A dedicated, specifically cleaned space is necessary for 
    sample processing.  A dedicated vehicle is preferable.  
    However, a specifically dedicated space within a vehicle might 
    suffice.  Such dedicated space should be separated from the 
    remaining space in the vehicle.  Also, use of a multi-purpose 
    vehicle will require a higher frequency of QC data to ensure 
    lack of contamination in processed samples.  Perhaps a small 
    camper trailer would be an acceptable alternative to either a 
    dedicated vehicle or dedicated space within a vehicle.  Ideas 
    for other alternatives should be communicated through the 
    QWTALK continuum.

Question 7:  Will I be able to obtain quick turnaround on QC 
    samples so I can determine if I am properly applying the 
    cleaning procedures and other quality-assurance procedures?

Answer:  The NWQL is developing a set of Custom Analyses 
    Schedules for analyzing deionized water blanks.  The specifics 
    are (1) a 2- to 3-week turnaround time; (2) detection levels 
    in the range of 0.2 to 0.4 ug/L, for 15-20 (actual number to 
    be determined) trace elements analyzed by ICP/MS, and (3) 
    analyses of nutrients, major ions, and additional trace 
    elements (such as arsenic, selenium, mercury) by existing 

    The ICP/MS method will be an extension of a proven method that 
    has reporting levels of 1 5g/L for 15 trace elements.  Thus, 
    the method will be experimental and the QC results will not be 
    reportable in Annual Data Reports or allowable in NWIS until 
    the method is approved.  The OWQ is extending the method 
    because QC data which have the same reporting level as 
    environmental data have very limited value.  The target is to 
    get the reporting level for the QC data below one-half the 
    reporting level of the environmental data.

    The NWQL will distribute a Technical Memorandum on this issue 
    in August.

Question 8:  Our office has a new system for producing deionized 
    water and initial QC results show that there is no trace- 
    element contamination.  Why should we pay the cost of buying 
    inorganic-free blank water (IFBW)?

Answer:  There are two reasons to buy the IFBW.  First, your past 
    analyses are probably at reporting levels of 1 ug/L for most 
    trace elements (unless custom analyses were run for specific 
    analytes).  As explained under question 7, this level provides 
    limited protection against the actual presence of 

    Second, what is the cost of consistent QC on the District 
    deionized water system, compared to the cost of buying IFBW? 
     The IFBW is QCed and virtually guaranteed contaminant free at
     all times.  In contrast, if a District system goes bad, and 
    QC samples are collected infrequently, all of the 
    environmental samples collected in the intervening time period
     could be contaminated.

Question 9:  You talk about collecting and interpreting QC data. 
    Where can I learn how to interpret the QC data?

Answer:  The Branch of Quality Assurance, in cooperation with the 
    OWQ, will conduct a Workshop on Interpretation, Management, 
    and Presentation of Water-Quality Quality-Control Data during 
    November 1-5, 1993, in Denver.  At least one person from each 
    District should attend and later provide training to the other 
    District personnel.

Question 10:  Given the lack of field method identification 
    capability in NWIS, if different programs and projects begin 
    using the protocol at different times, future data users will 
    not know what method was used to collect recorded data.  
    Therefore, shouldn't all programs and projects begin using the 
    protocol at the same time?

Answer:  The transition to using the protocol represents a 
    difficult cultural change in how the Division conducts water-
    quality work.  Some programs and projects are anxious to use 
    the protocol now (a few already are).  Others might not be 
    ready for a year or more.  We are making the stated decision 
    so as many projects as possible can begin producing good- 
    quality data as quickly as possible.  To continue to produce 
    poor-quality data or even no data (when good-quality data 
    could be produced) is unacceptable.

Question 11:  What is the bottom line on whether we need to 
    follow all aspects of the protocol at all times?

Answer:  There are four points to consider.  First, the protocol 
    is provided as a guide to improving WRD's data-collection 
    efforts.  The key messages of the protocol are: (1) inorganic 
    samples can be contaminated, but sources of contamination can 
    be reduced through proper planning, use of tested equipment/ 
    supplies, proper cleaning, and specified quality-assurance 
    measures; and (2) collection of adequate QC data will identify 
    whether problems exist (suggested QC in the protocol should be 
    considered only a minimum guide to what may be needed to 
    identify the problems).  

    Second, the OWQ has data to prove that the protocol works for 
    trace elements, nutrients, and major ions, provided the 
    recommended equipment and supplies are used and all cleaning 
    and other quality-assurance procedures are appropriately 

    Third, based on cost, time, and availability of personnel, 
    situations may arise wherein it would be desirable to ignore 
    all or parts of the protocol.  However, this creates the risk 
    of contamination and bad data.  When shortcuts are taken, the 
    project must develop the appropriate types and amounts of QC 
    data to ensure that the changes do not produce contamination 
    in the results.  The types and amounts of QC data required 
    will depend on (1) the overall objects of the program or 
    project and (2) the data-quality objectives of each specific 
    data-collection activity.

    Fourth, each deviation from the protocol constitutes an 
    experiment the District or project must conduct to ensure lack 
    of contamination.  Furthermore, the efficacy of deviating in 
    multiple procedures or steps constitutes a multistep 
    experiment because, if the QC data show contamination, each 
    specific deviation will need to be evaluated separately to 
    find the source(s).

    As more experience is gained using the protocol, modifications
    will be made.  However,  for now consider that parts of the 
    protocol (e.g., cleaning methods, equipment, supplies) are 
    relatively easy to apply and much less expensive than poor-
    quality data. 

The OWQ is implementing the protocol on October 1, 1993, in order 
to provide the earliest possible benefit to the Division.  
Employment of the procedures contained in the protocol is an 
important step to improving data quality for WRD's mission.  A 
preliminary list of equipment and supplies discussed in the 
protocol is attached for informational purposes.  Recall that the 
OWQ will issue a Technical Memorandum to present the protocol as 
Division policy.

                               David A. Rickert
                               Chief, Office of Water Quality


This memorandum does not supersede any previous Office of Water 
Quality Technical Memorandum.

Key Words:  field procedures, inorganic, trace elements

Distribution:  A, B, S, FO, PO

                        ATTACHMENT 1

                      SURFACE-WATER SAMPLING

Non-contaminating sampler (DH-81, D-77, Frame)
      with polyethylene or Teflon bottle or bag, cap, and nozzle
Churn splitter (8 or 14 L) (Bel-Art Products)
Carrier for churn splitter (plastic container with lid)
Processing and preservation chambers (cpvc or pvc)
Plastic bags, clear for use in chambers
Capsule filters, Supor (Gelman 12175/12176) OR
      142mm plate filter and MFS cellulose acetate filters
      Non-metallic forceps for handling filters
Peristaltic pump for filtration
Pump tubing (silicon or Teflon)
Non-powdered vinyl gloves, disposable
Concentrated hydrochloric acid (HCl), Baker Instra-Analyzed
Deionized water (DIW)
Tap water
Wash basins, clear plastic
Non-metallic brushes
Wash bottles, safety labeled, for dilute HCl, DIW
Sealable plastic bags (Ziplock) for storing cleaned equipment
Inorganic-free blank water

DH-81 and D-77 samplers are currently available from the Waterways
Experiment Station in Vicksburg, Miss. (Contact Wayne O'Neal, 
601-634-2721).  Plans for the frame sampler will be available from 
the Office of Water Quality.  All other equipment and supplies, 
except the water, will be available by October 1, 1993, from the 
Quality Water Services Unit in Ocala, Fl.

Plans for processing and preservation chambers are available 
as Framemaker files.  They can be obtained by:

login: anonymous
password:  your userid
cd pub