Ammonia Matrix Spike Recovery: QC for Wastewater Reporting
Defending the Batch: Recovery and Wastewater Reporting
A finished ammonia batch still has to show that the number is in control before it is reported. Two fortified checks do that work. The laboratory fortified blank asks whether the method recovers ammonia in reagent water. The ammonia matrix spike recovery asks whether this wastewater sample does.
Section 9 of Timberline Method Ammonia-001 is explicit about a miss. The batch is out of control. The laboratory finds the cause, writes it down, corrects it, and repeats the batch. The method does not offer a qualifier code as a way to report the failed batch.
The Method-Specific QC Framework
The windows below belong to gas diffusion with conductivity detection, as written in Section 9 of Timberline Method Ammonia-001. They are not universal EPA percentages. A permit, a different approved method, or a later revision can set other limits. The document that governs the report controls.
EPA Method 350.1 is the phenate method. Its fortified-matrix window is 90–110% in that method. Do not apply that window to this gas-diffusion procedure, and do not apply section 9’s windows to a phenate run.
The checks also assume a valid calibration. The method allows a linear or quadratic fit and requires a correlation coefficient (R) of 0.995 or higher. Standards are prepared in reagent water and made fresh weekly. Stock preparation and reporting as N are in How to Prepare Ammonia Calibration Standards. A spike calculated against a curve that fails that fit does not meet the method’s starting condition.
Laboratory Fortified Blank (LFB): System Performance Check
Section 9.5 defines the laboratory fortified blank. Reagent water is spiked with ammonia at a concentration between 10% and 50% of the dynamic range in use, and the fortified blank is analyzed at least once per batch. The control window in this method is 87–104% recovery.
The fortified blank is the clean-water check. It does not describe the wastewater matrix. If recovery falls outside 87–104%, section 9.5 does not publish a diagnostic checklist. It says the batch is out of control: find the cause, record it, take corrective action, and repeat the batch. Sample results from that batch are not reported until a batch is in control. Do not treat pump tubing, buffer age, or caustic strength as method-required causes. They are places an operator may look. They are not the acceptance rule.
Matrix Spike (LFM/LFMD): Assessing Sample Interference
Section 9.6 is the sample check. A duplicate pair of ammonia samples is spiked with ammonia at 10–50% of the same dynamic range. The method also allows a Kjeldahl sample as that matrix. Kjeldahl measurement itself is covered in Streamlining TKN and TN Analysis. This article does not turn the spike into a TKN procedure.
Section 9.6 sets two limits:
- Recovery: 84–115%.
- Relative percent difference between the duplicate pair: no more than 20%.
Outside either limit, the corrective action is the same as for the fortified blank. The batch is out of control and is repeated after the cause is recorded and corrected.
Interpreting Failures: Gas Diffusion and Matrix Tolerance
A fortified blank can pass while the matrix spike fails. That combination means the clean-water check recovered ammonia and this sample did not. It does not identify the cause. Do not attribute the miss to turbidity, color, particulates, heavy metals, or oxygen demand. The method does not.
Laboratories choose gas diffusion because color and particulates interfere with optical ammonia methods. The sample is mixed with caustic so ammonium becomes dissolved ammonia gas. The gas crosses a tubular hydrophobic membrane into the absorbing buffer, and conductivity rises in proportion to the ammonium that crossed. The TL-2800 ASX datasheet describes that design for turbid, colored, and particulate-laden samples. Tolerance is not immunity. The method already warns that interfering volatile conductive compounds can be present. A failed spike is still a failed spike until the batch is repeated in control.
A spike that passes does not excuse a failed reagent blank. The blank rule is in When an Ammonia Reagent Blank Means the Run Stops.
Corrective Actions for Out-of-Control Batches
When fortified-blank recovery, matrix-spike recovery, or the duplicate relative percent difference falls outside section 9.5 or 9.6, reporting stops. The method’s sequence is root cause, corrective action, a written record, and a repeat of the batch. It does not provide J-flags, estimated results, or another qualified-reporting path around those windows.
Section 9 does not list the hardware checks that will turn out to be the cause. Those depend on the batch. What the method does require is that the failed batch is not the batch that gets reported.
Dynamic Range and Spike Level Selection
The 10–50% spike is 10–50% of the dynamic range the laboratory is calibrating, not a single universal concentration. Table 2 of Timberline Method Ammonia-001 gives the published ranges. The high-sensitivity 10-inch membrane is 0.020–2 mg/L ammonia-N. The standard 4-inch membrane is 0.050–500 mg/L ammonia-N.
A laboratory may calibrate a shorter curve than the full Table 2 span. The fortified result still has to land inside the curve actually in use. If that curve stops at 50 mg/L and the sample is already 45 mg/L, a 20 mg/L spike carries the total past the top standard. That spike does not show recovery on the calibrated range. Dilute before spiking, or choose a spike that stays inside the curve. The 50 mg/L figure in that example is a shorter working curve, not the top of Table 2.
Ongoing Demonstration of Capability and Regulatory Scope
Section 9.7 charts laboratory-fortified-blank recovery as the ongoing demonstration of capability and compares it with the laboratory’s control-chart limits. The method extract cuts off in the middle of the corrective-action sentence for that chart. Do not invent the rest. Initial demonstration of capability and the wider validation file are in Validating Ammonia Analyzer Methods for Regulatory Compliance.
A zinc reduction cartridge lets the same detector support a total inorganic nitrogen reading by reducing nitrate and nitrite to ammonium. The instrument does not speciate nitrate from nitrite. EPA approval of this method is for ammonia. A passing ammonia fortified blank or matrix spike is not nitrate-recovery evidence and is not Clean Water Act approval for nitrate, nitrite, or total inorganic nitrogen. Those uses stay informational or process-monitoring data unless a documented variance says otherwise. Listed ammonia methods are in 40 CFR 136 Table IB for Ammonia.
Technical Resources and Support
Laboratories whose phenate or electrode method cannot hold a matrix spike on colored or particulate effluent can review the ammonia analyzer as the gas-diffusion option. The spike study is still required on that instrument. It is not waived by the method’s matrix tolerance.
Laboratories that already run the instrument and need the fortified-blank and matrix-spike sequence written into an SOP, or that cannot get a batch back inside section 9, should use Timberline training and support. The first response to a failed spike is to repeat the batch in control, not to replace the analyzer.
References
- Method 350.1, Revision 2.0: Determination of Ammonia Nitrogen by Semi-Automated Colorimetry. U.S. Environmental Protection Agency (1993). https://www.epa.gov/sites/default/files/2015-08/documents/method_350-1_1993.pdf
- Timberline Ammonia-001: Determination of Inorganic Ammonia by Continuous Flow Gas Diffusion and Conductivity Cell Analysis. Timberline Instruments (2011). https://www.timberlineinstruments.com/wp-content/uploads/2019/07/Ammonia-EPA-Method_6-24-2011_Final_All.pdf
- TL-2800 ASX datasheet (DSM-280000-AA). Timberline Instruments. https://www.timberlineinstruments.com/wp-content/uploads/2025/03/DSM-280000-AA-TL2800-ASX.pdf