14.3 Laboratory QA/QC, Chain of Custody & Compliance Records

Foundational Principles: Quality Assurance vs. Quality Control

Drinking water laboratories generate data that dictate multi-million-dollar treatment adjustments, chemical dosing regimens, and emergency public health advisories. Analytical results must be technically defensible, repeatable, and legally unassailable. Regulatory laboratories operate under a dual framework: Quality Assurance (QA) and Quality Control (QC).

The Quality Assurance (QA) Umbrella

Quality Assurance (QA) is the overarching, managerial program that guarantees analytical data are of known and documented quality. QA encompasses all policies, administrative procedures, operational structures, and facility guidelines:

• Formulation of the Quality Assurance Manual (QAM), which establishes organizational hierarchy, data verification pathways, and corrective action protocols.
• Maintenance of certified Standard Operating Procedures (SOPs) for every analytical method performed within the facility.
• Mandatory analyst training, documented Demonstrations of Capability (DOC), and semi-annual blind Proficiency Testing (PT) audits administered by third-party accrediting bodies.
• Documented preventative maintenance schedules, instrument calibration records, thermometer calibrations traceable to National Institute of Standards and Technology (NIST), and climate-controlled laboratory logs.

Quality Control (QC) Operational Mechanics

Quality Control (QC) consists of the specific tactical procedures and laboratory checks executed on a daily, batch-by-batch basis to measure the performance of analytical equipment, chemical reagents, and technicians. QC provides the numerical data used to quantify accuracy (how close an analytical result is to the true value) and precision (how repeatable results are when analyzing identical aliquots).

[ Quality Assurance (QA) - Overarching Management System ]
├── Personnel Certification & Training (DOC)
├── SOPs, Instrument Logs & NIST Traceability
└── Quality Control (QC) - Daily Batch Operations
    ├── Accuracy Assessment: Matrix Spikes (MS/MSD), Calibration Verification (CCV)
    ├── Precision Assessment: Sample Duplicates (RPD)
    └── Contamination Verification: Method Blanks, Trip Blanks

---

Quality Control Sample Classifications & Mathematical Metrics

To ensure analytical batches satisfy regulatory standards, analysts process specialized QC samples alongside raw environmental samples:

1. Blanks (Assessing Background Contamination)

• Method Blank (Reagent Blank / Laboratory Blank): An aliquot of analyte-free, high-purity laboratory water (ASTM Type I or II) that is processed through the entire analytical sequence, including all sample digestion, chemical preservation, reagents, and filtration steps. The method blank detects contamination originating from impure chemical reagents, glassware leaching, contaminated apparatus, or ambient airborne dust. Acceptance Standard: Must be below the Method Detection Limit (MDL) or Reporting Limit (RL).
• Field Blank: Reagent water filled in the laboratory, transported to the field sampling station in an empty container, opened and transferred to a standard collection bottle at the site, and handled identically to an environmental sample. Measures contamination introduced by ambient field dust, vehicle exhaust, or sampling technique.
• Trip Blank: Sealed vials of organic-free reagent water prepared in the laboratory that accompany empty sample bottles into the field in the transport cooler, remain unopened throughout the sampling event, and return to the laboratory for analysis. Trip blanks are mandatory when sampling for Volatile Organic Compounds (VOCs) to determine if airborne VOC vapors permeated container septa during vehicular transit or storage.

2. Standards (Assessing Calibration Accuracy)

• Initial Calibration Verification (ICV): A standard solution prepared from an independent chemical source (a secondary manufacturer or different chemical lot from the calibration standards). Analyzed immediately following the initial multi-point calibration curve to confirm instrument calibration validity.
• Continuing Calibration Verification (CCV): A calibration standard analyzed at regular intervals throughout the analytical run (typically every 10 to 20 samples and at the conclusion of the analytical sequence). Verifies that instrument sensitivity has not drifted. Acceptance Standard: Typically requires 90% to 110% recovery of the true standard concentration. If a CCV fails, all environmental samples analyzed subsequent to the last passing CCV must be re-analyzed.

3. Matrix Spikes (Assessing Analytical Accuracy and Matrix Interference)

A Matrix Spike (MS) is an aliquot of an actual field drinking water sample to which a known concentration of target analyte is deliberately added prior to digestion or analytical processing. Processing the spiked field sample alongside an unspiked sample reveals whether physical or chemical properties of the raw water (such as extreme hardness, turbidity, organic acids, or high total dissolved solids) suppress or artificially enhance the instrument signal—a phenomenon known as matrix interference.

Analytical accuracy is quantified as Percent Recovery (% Recovery):

$\text{Spike Recovery } (\%) = \left[ \frac{C_{\text{spiked}} - C_{\text{unspiked}}}{C_{\text{spike added}}} \right] \times 100$

Where:

• $C_{\text{spiked}}$ = Total measured concentration in the spiked field sample aliquot
• $C_{\text{unspiked}}$ = Measured concentration in the original unspiked field sample
• $C_{\text{spike added}}$ = Theoretical concentration of analyte added to the sample aliquot

Acceptance Standard: Typically 80% to 120% or 70% to 130% recovery, depending on EPA method specifications.

4. Duplicates (Assessing Analytical Precision)

• Laboratory Duplicates: Two separate aliquots drawn from a single, thoroughly homogenized field sample bottle and carried through independent analytical processing. Assesses the repeatability and precision of the laboratory's analytical method and technician technique.
• Field Duplicates: Two independent sample containers filled sequentially from the identical field tap and analyzed to measure combined field sampling variation and laboratory precision.

Precision between duplicate analyses ($X_1$ and $X_2$) is mathematically evaluated as the Relative Percent Difference (RPD):

$\text{RPD } (\%) = \left[ \frac{|X_1 - X_2|}{\frac{X_1 + X_2}{2}} \right] \times 100 = \left[ \frac{2 \cdot |X_1 - X_2|}{X_1 + X_2} \right] \times 100$

Where:

• $X_1$ = Measured concentration of the primary sample
• $X_2$ = Measured concentration of the duplicate sample

Acceptance Standard: For environmental samples containing concentrations well above the reporting limit, RPD must typically remain $\le 10%$ to $20%$.

---

Statistical Process Control: Shewhart Control Charts

To detect systematic drift, analytical bias, and random mechanical errors before they compromise compliance results, water laboratories maintain Shewhart Control Charts. A control chart plots consecutive QC measurements (such as daily percent recovery of a check standard or matrix spike) on the vertical axis against analytical batch sequence or time on the horizontal axis.

+-------------------------------------------------------------+ Upper Control / Action Limit (+3s)
|                                                             
+- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -+ Upper Warning Limit (+2s)
|                                                             
|======================== Mean (μ) ===========================| Center Line
|                                                             
+- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -+ Lower Warning Limit (-2s)
|                                                             
+-------------------------------------------------------------+ Lower Control / Action Limit (-3s)

Statistical Boundaries and Standard Deviations

Control charts are constructed from historical baseline data (typically 20 to 30 consecutive, independent QC runs under stable conditions) using the mean ($\mu$) and standard deviation ($s$):

1. Center Line (Mean, $\mu$): Represents the average analytical recovery or target value.
2. Warning Limits (WL): Established at $\pm 2\text{ Standard Deviations}$ ($\pm 2s$) from the mean. In a Gaussian normal distribution, 95.45% of all random data points naturally fall within $\pm 2s$. A single point crossing a warning limit alerts the analyst to possible degradation, but does not invalidate the run unless consecutive points demonstrate drift.
3. Control / Action Limits (CL): Established at $\pm 3\text{ Standard Deviations}$ ($\pm 3s$) from the mean. In a normal distribution, 99.73% of all data points fall within $\pm 3s$. The probability of a valid run falling outside $\pm 3s$ strictly by chance is less than 0.3%.

Out-of-Control Operating Rules

An analytical sequence is officially designated "out-of-control"—requiring immediate termination of analysis, formal investigation, and complete re-analysis of all samples in that batch—if any of the following statistical rules are violated:

• Single Excursion Beyond Action Limits: Any single data point falls outside the $\pm 3s$ Action Limits.
• Two Consecutive Points Beyond Warning Limits: Two consecutive data points fall outside the $\pm 2s$ Warning Limits on the same side of the mean.
• Systematic Shift (Bias): Seven consecutive data points fall entirely on one side of the center line, indicating systematic calibration error or reagent contamination.
• Systematic Trend (Drift): Six consecutive data points steadily increase or steadily decrease, indicating instrument detector decay, lamp aging, or reagent evaporation.

---

Chain of Custody (COC) Legal & Technical Architecture

Drinking water compliance samples can become central evidence in state administrative hearings, civil litigation, or federal enforcement actions under the SDWA. A Chain of Custody (COC) is a legally binding document that establishes an unbroken, chronological record of physical custody, transfer, and disposition of a sample from the precise moment of field collection to final laboratory destruction.

Mandatory Chain of Custody Information

A legally complete COC form must contain:

1. Unique Sample Identification Number: Cross-referenced with the waterproof label on the bottle.
2. Public Water System (PWS) Details: PWS name, 7-digit PWSID number, facility address, and phone number.
3. Collection Specifics: Exact physical tap location (e.g., "Treatment Plant Entry Point #1" or "104 Elm Street Outside Hose Bibb"), sample collection date, and exact military collection time.
4. Sample Matrix & Type: Potable finished water, raw groundwater, surface water, or distribution network; designated as either Grab or Composite.
5. Field Preservation Documented: Specific chemical preservatives added ($HNO_3$, $H_2SO_4$, $NaOH$, or $Na_2S_2O_3$) and field temperature verification.
6. Requested Analytical Parameters: Specific target compounds and requested EPA Method numbers.
7. Sampler Certification: Sampler's printed name and physical signature.
8. Custody Signatures: Relinquished By and Received By lines documenting the exact date, military time, and institutional affiliation of every transfer. The presence of intact, serialized tamper-evident custody seals across cooler lids is logged.
9. Cooler Receipt Verification: Laboratory intake staff must measure and record the internal temperature of the transport cooler immediately upon arrival (mandated at $\le 6^\circ\text{C}$, but above freezing, typically $4 \pm 2^\circ\text{C}$ on wet ice). Frozen or warm samples are flagged as compromised.

---

SDWA Statutory Record Retention Mandates

The EPA enforces explicit record retention timelines under 40 CFR § 141.33. Public water system operators must maintain permanent records on site or readily retrievable for state primacy agency review:

[ Safe Drinking Water Act Record Retention Hierarchy ]
├── 5 Years  --> Bacteriological Analysis Records (Colilert, MF, HPC)
├── 5 Years  --> Variances or Exemptions (Following Expiration Date)
├── 10 Years --> Chemical Compliance Records (Inorganics, Organics, Metals, Radiochemicals)
├── 10 Years --> Sanitary Surveys, State Written Audits & Corrective Action Letters
└── 12 Years --> Lead and Copper Rule (LCR) Monitoring & Compliance Records

• Bacteriological Records (Minimum 5 Years): All records of microbiological analyses (Total Coliform, E. coli, HPC) must be retained for at least five years. The record must include date, place, and exact time of sampling; identifier of the sampler; whether the sample was routine, check, raw, or special; date of analysis; laboratory and analyst; analytical method; and test results.
• Chemical Analysis Records (Minimum 10 Years): Records of chemical compliance analyses (nitrate, nitrite, fluoride, synthetic organics, volatile organics, disinfection byproducts, secondary metals) must be retained for at least ten years.
• Lead and Copper Rule Records (Minimum 12 Years): All documentation relating to the Lead and Copper Rule—including tap monitoring results, 90th percentile calculations, customer tier documentation, corrosion control studies, and lead service line inventories—must be retained for at least twelve years.
• Sanitary Survey Reports & Written Communications (Minimum 10 Years): Official written reports, comprehensive performance evaluations (CPE), sanitary surveys conducted by the state, and formal letters regarding plant deficiencies or compliance violations must be kept for at least ten years.
• Variances or Exemptions (Minimum 5 Years Following Expiration): Records concerning a variance or exemption granted to the utility must be retained for at least five years following the date that the variance or exemption expired.

---

Comparative Quality Management Tables

Table 14.3.1: Laboratory Quality Control Samples & Acceptance Specifications

Table 14.3.2: Safe Drinking Water Act Compliance Record Retention Mandates