#269 2006f

Page 1 of 5
#269
Severn Trent Savannah Laboratory, Inc.

On-Site Inspection Report

LABORATORY NAME: Severn Trent Savannah Laboratory, Inc.

ADDRESS: P.O. Box 13548
5102 La Roche Ave.
Savannah, GA 31404

CERTIFICATE NO: 269

DATE OF INSPECTION: November 06, 07 & 08, 2006.

TYPE OF INSPECTION: Maintenance

EVALUATOR: Mr. Jeffrey R. Adams, Mr. David Livingston and Mr. Gary Francies

LOCAL PERSON(S) CONTACTED: Ms. Andrea Teal, Ms. Ammie Martin, Mr. Myron Young and Mr. Ben
Gulizia.

I. INTRODUCTION:

This laboratory was inspected to verify its compliance with the requirements of 15A NCAC 2H .0800 for the
analysis of environmental samples.

II. GENERAL COMMENTS:

NOTE: Whenever a certified commercial laboratory refers or subcontracts samples to another certified
laboratory for analyses, the referring laboratory must supply the date and time samples were collected to
insure holding times are met. Subcontracted samples must clearly indicate the State of North Carolina as
the collection site on all record transcripts. Laboratories may subcontract sample fractions, extracts,
leachates and other sample preparation products provided that all Rules and requirements of 15A NCAC
02H .0800 are documented. The initial client requesting the analyses must receive the original or a copy of
the report made by the laboratory that performs the analyses.
Ref: 15A NCAC 02H .0805 (e) 2.

NOTE: Records which are stored only on electronic media must be maintained and supported in the
laboratory by all hardware and software necessary for immediate data retrieval and review. Ref.15A
NCAC 2H .0805 (a) (7) (G).

NOTE: At any time a laboratory receives samples which do not meet sample collection, holding time, or
preservation requirements, the laboratory must notify the sample collector or client and secure another
sample if possible. If another sample cannot be secured the original sample may be analyzed, but the
results reported must be qualified with the nature of the infraction(s) and the laboratory must notify the
State Laboratory about the infraction(s). The notification must include a statement indicating corrective
action taken to prevent the problem for future samples. Ref. 15A NCAC 2H .0805 (a) (7) (N).

III. DEFICIENCIES, REQUIREMENTS, COMMENTS AND RECOMMENDATIONS:

Quality Control:

KONELAB PARAMETERS and TOC INSTRUMENT SOFTWARE

Page 2 of 5
#269
Severn Trent Savannah Laboratory, Inc.
RECOMMENDATION: The software for the Konelab and TOC instruments flag any Blank value that is
not exactly zero as “QC rule violation”. It is strongly recommended that this either be fixed or if that is
not possible then manually mark the printouts “OK” and initial during the QC data review.

A. COMMENT: The laboratory is not making corrections to the data as required.

REQUIREMENT: If a mistake is made in data recording, the mistake must be removed by making a
single line through the incorrect entry, recording the correct entry beside mistake, and initialing and
dating the error. Ref: SW 846 Sept. 1994 Test Methods for Evaluating Wastes, Chapter One pg. 13

WALKTHROUGH:

INORGANICS:

OVENS

1. DEFICIENCY: The temperature is not documented for the TOC, TSS, TR, or TDS ovens.

NOTE: This documentation was inadvertently discontinued in the conversion to the new LIMS system.

REQUIREMENT: Each day an incubator, oven, waterbath or refrigerator is used, the temperature must
be checked, recorded, and initialed. Ref: 15A NCAC 2H .0805 (a) (7) (J).

B. COMMENT: The time in and out of the TDS oven is not documented.

REQUIREMENT: In order to document the method required one hour drying time of 180 ± 20 C the time
evaporated samples are placed in the 1800 C oven and the time removed must be documented. Ref:
Standard Methods, 20th Edition – Method 2540 C. 3d.

AMMONIA NITROGEN AUTOMATED and FLUORIDE - ELECTRODE

C. COMMENT: No distillation studies were performed.

REQUIREMENT: All samples must either be distilled or a distillation study must be on file to show the
distillation step is not needed. Commercial Laboratories shall analyze initially at least six samples distilled
and undistilled, but at least two samples from all waste types. Analyze one sample distilled and undistilled
yearly from all wastes types. Additional samples must be analyzed comparatively if the above sample
results do not indicate approximately the same values for distilled and undistilled samples. Ref: Federal
Register, January 31, 1994; 40 CFR 136; Footnote 6 and NCLC Policy.

NOTE: Currently the lab is certified for both the Fluoride electrode and IC methods and samples are
analyzed by both methods. Since no distillation is required for the IC method it is recommended that
certification for the electrode method be discontinued. A distillation study will have to be performed for
ammonia nitrogen. Since no distillation comparison data was on file the laboratory will be required to
perform the initial requirement of six samples. Please submit this data upon completion.

TOTAL SUSPENDED RESIDUE

D. COMMENT: Sufficient sample volume is not always analyzed to obtain the method required weight
gain.

NOTE: A maximum volume of 100 ml is generally used. The lab’s reporting limit is currently 5mg/l. This
reporting limit must be changed to 10 mg/l if a maximum sample volume of 100 mls is to be continued.
Page 3 of 5
#269
Severn Trent Savannah Laboratory, Inc.

In instances where the weight gain is less than the required 1.0 mg, the value must be reported as less
than the appropriate value based upon the volume used. It is strongly recommended that a minimum
volume of 200 mls be filtered when possible in order to maintain the reporting limit of 5 mg/l.

REQUIREMENT: When analyzing suspended residue, sufficient sample must be filtered to yield at
least 1.0 mg of residue. Ref: EPA-600/4-79-020 Methods Manual, March 1983 - Method 160.2–2.IV.

CYANIDE

RECOMMENDATION: All calibration curves are second order. It is recommended that the possibility of
using linear calibration be considered.

BOD

RECOMMENDATION: It is strongly recommended that a seeded blank be analyzed with each set of
samples. Its value should be compared with the calculated seed correction value to help judge the
consistency of the seeding material and analyst technique. In general the difference between the
seeded blank concentration and the calculated seed correction factor should be no more than
approximately 0.2 mg/l.

E. COMMENT: Samples are not being properly checked for chlorine and neutralized with sodium sulfite
if necessary.

NOTE: After prescreening with DPD, any samples containing chlorine are treated with solid sodium
sulfite. This often will result in an excess amount of sodium sulfite creating of an additional oxygen
demand.

REQUIREMENT: Samples containing residual chlorine compounds must be dechlorinated prior to
analysis and the samples seeded. This procedure must be documented on the worksheet. The chlorine
must be neutralized according to the titration procedure outlined in the method. Residual chlorine will
inhibit oxygen depletion due to killing or suppressing the bacteria and excess sodium sulfite will create
an oxygen demand that will result in false high oxygen depletion. Ref: Standard Methods, 20th Edition -
Method 5210 B. 4.e 2).

F. COMMENT: The laboratory has been using a true value of 200 mg/L for the glucose-glutamic acid
(GGA) standard. The SOP indicates the concentration is 198 mg/l. An acceptance range for the GGA for
CBOD of 81-119% is used.

REQUIREMENT: For both BOD and CBOD, the concentration of the glucose-glutamic acid standard is 198
± 30.5 mg/l. Ref: Standard Methods, 20th Edition - Method 5210 B.6.

SETTLEABLE RESIDUE:

RECOMMENDATION: It is strongly recommended that the start/end times be documented to indicate that
the required one hour is used. Currently one hour is used but not documented.

PHENOL

1. DEFICIENCY: No pH adjustment is performed or documented.
NOTE: The SOP indicates that the samples must be adjusted to pH 10 ± 0.2 S.U.

Page 4 of 5
#269
Severn Trent Savannah Laboratory, Inc.
REQUIREMENT: The approved method is cited in Standard Methods 14th Edition. The colorimetric
reaction is conducted at pH 10 ± 0.2 S.U. (The correct pH is footnoted in subsequent editions of
Standard Methods). Ref: Federal Register, 40 CFR-136.3, October 23, 2002.

TOC:

RECOMMENDATION: It is strongly recommended that a bottle blank be analyzed per lot of bottles
prior to their use. Currently the laboratory is using certified bottles.

FECAL COLIFORM

G. COMMENT: No comparison test is conducted before a new lot of consumables are put into use.

NOTE: At the present time the laboratory is performing a negative check only. A handout describing
this procedure was provided.

REQUIREMENT: It is required that when a new lot of culture medium, pads, or membrane filters is to
be used, a comparison of the current lot in use, (reference lot) against the new lot (test lot), be made.
As a minimum, make single analyses on five positive samples. Ref: Standard Methods, 18th Edition -
Method 9020 B.3d. (page 9-7).

FECAL and TOTAL COLIFORM

H. COMMENT: Only an initial dilution water blank is analyzed.

REQUIREMENT: It is required that a negative control (blank) be analyzed with each series of samples
using buffered water and the medium batch at the start of the test series, following every tenth sample,
and the end of the test series. Ref: Standard Methods, 20th Edition - Method 9222 B. 5c. and
Microbiological Methods for Monitoring the Environment part IV. Section C. 1.3.

I. COMMENT: Only one dilution is analyzed for each sample.

REQUIREMENT: It is required that multiple dilutions be used in order to better ensure that the required
colonies (20-60 Fecal and 20-80 Total) are obtained. Ref: Standard Methods, 20th Edition – Method 9222
D. 2a. and 9222 B. 5a. See Tables 9222:I and Table 9222:III for suggested sample volumes.

METALS:

EPA METHODS: 200.7, 6010B; 200.8/6020

EPA METHOD 3030C:

J. COMMENT: The laboratory stated that it may not always be performing 3030C when performing metals
analysis on NC groundwater samples if not specifically noted on login.

REQUIREMENT: Groundwater samples from the state of North Carolina must be digested by Method
3030 C and documented to have been collected and digested within 72 hours. Ref: Groundwater Section
Memo February 22, 1993. Samples are not to be evaporated down as indicated in Method 3010A,
samples should be treated as described in Method 3030 C.

RECOMMENDATION: It is highly recommended that the laboratory construct a list of NC requirements
and require that all NC samples be flagged with the appropriate handling and analytical requirements.

Page 5 of 5
#269
Severn Trent Savannah Laboratory, Inc.
K. COMMENT: The lab is using an acceptance criterion that the blanks must be less than the reporting
limit standard (PQL).

REQUIREMENT: NCLC policy requires that the concentration of the blank must not exceed 50% of the
LOQ/MQL/PQL. The lab must obtain a signal in which the constituent concentration in reagent water is
2(1.645)s above the mean of blank analyses (Lower Limit of Detection). The constituent concentration
that produces a signal sufficiently greater than the blank, that it can be detected within specified limits,
by good laboratories, during routine operating conditions, is the Limit of Quantitation. Typically it is the
concentration that produces a signal 10s above the reagent water blank signal. In most cases, this
requires an absorbance value greater than 0.005 units, preferably greater than 0.010 units. Ref:
Standard Methods, 20th Edition - Method 1010 C. and 1030 C.

NOTE: The laboratory was unaware of the < 50% of PQL rule. With the exception of only a very few
elements, most blanks exceeded the < <50% requirement. The 2006 laboratory acceptance criterion
was much tighter than in the previous year.

ORGANICS:

L. COMMENT: The laboratory does not routinely check all organic samples for Residual Chlorine
content.
REQUIREMENT: For any sample that requires the removal of residual chlorine, (8260,8270,8021,
TKN, Ammonia, BOD, etc.) the laboratory must check to ensure that the chlorine has been removed.
The residual chlorine check must be performed to verify that the residual chlorine level is below 0.5
mg/L. Commercially available starch iodide paper can be used to perform this determination.

NOTE: During the audit, the QA coordinator submitted an “Amendment to Certification Application”
requesting the dropping and adding of several parameters. That process has since been completed and a
new updated certificate of those changes has been submitted to the laboratory.

IV. PAPER TRAIL INVESTIGATION:

The inspectors reviewed several data packs from December 05, 2005 to September 09, 2006 for most
parameters listed on the laboratory’s certificate. This consisted of comparing data with the values obtained
on laboratory bench worksheets (for both water and soil analysis) including calibrations, quality control
charts, surrogate and duplicate recoveries and statistical calculations performed on duplicate analysis
monitoring precision and accuracy. With the exception of the above statements, the inspectors found no
computation errors on the data packs reviewed and data appeared accurate and easily accessible to the
inspector.

V. CONCLUSIONS:

Correcting the above deficiencies and comments and implementing the recommendations should help
this laboratory produce quality data and meet certification requirements. Again, the inspectors would
like to thank the laboratory staff for their great co-operation in the inspection and data review process.

Please respond to all numbered Deficiencies and lettered Comments cited in this report.

Report prepared by: Jeffrey R. Adams Date: December 4, 2006
Report Reviewed by: Gary Francies Date: December 6, 2006
Report Reviewed by: David G. Livingston Date: December 7, 2006