GROUNDWATER MONITORING BASICS
What a statistically significant increase is, and what happens after one
30 September 2026 · about a six minute read
Part of a series. If you are new to the subject, start with what RCRA groundwater monitoring is, and why it needs statistics.
A statistically significant increase, usually shortened to SSI, is the finding that starts almost everything else in landfill groundwater monitoring. It has a precise meaning in the federal rule, and it is not the same thing as a concentration going up. This article explains what an SSI is, how one is determined, and what the rule requires once one is found.
It covers municipal solid waste landfills regulated under RCRA Subtitle D, the part of the Resource Conservation and Recovery Act that governs them, and the groundwater rules in 40 CFR Part 258, Subpart E. States with approved programs can set different schedules and add requirements, and a facility's permit may say more, so check what applies to your site.
Detection monitoring, in one paragraph
A landfill starts in detection monitoring. At least twice a year, it samples its wells and has them analyzed for the constituents in Appendix I of Part 258, or an alternative list its state has approved. Appendix I is a list of metals and volatile organic compounds, the ones most likely to show that the landfill is leaking. The results from the compliance wells at the edge of the landfill are then compared with background, the groundwater the landfill has not affected.
What makes an increase statistically significant
For each constituent at each compliance well, the owner or operator must decide whether there is a statistically significant increase over background. The rule names the statistical methods that may be used, and every one of them compares the compliance well result with background:
- an analysis of variance, comparing each well's mean or median with background
- a tolerance or prediction interval computed from background, with each compliance well result compared with its upper limit
- a control chart
- another method, if it meets the same performance standards
So an SSI is the result of a test, not an observation. A concentration can go up without being an SSI, because groundwater varies naturally and a single higher result can fall well within what background says is normal. And a concentration can be an SSI without being very high, if background shows that the constituent is normally steady and low at that well.
Why a single high result is not the end of the question
A landfill with ten compliance wells and fifteen constituents runs one hundred and fifty comparisons every event. Even when nothing is wrong, some of those comparisons will come out high by chance. These are false positives, and with enough comparisons they are close to certain unless the statistical program is designed to control them. The Unified Guidance notes that across all of a site's wells and constituents, the false positive rate can be much greater than the rate for any single test.
The usual way to control them with prediction limits is retesting, also called verification resampling. The Unified Guidance, EPA's guidance on the statistics these rules require, describes it this way: when an initial result suggests contamination, one or more additional samples are collected and tested to see whether the first result was accurate. Resamples are taken only at wells where the initial result exceeded the limit. The guidance describes several schemes. In a "1-of-2" scheme, a well passes if its initial result is within the limit. If it is not, one resample is taken, and the well passes if the resample is within the limit. Which scheme a facility uses belongs to its statistical program, and in some states it is written into the facility's permit.
What the rule requires after an SSI
Once an SSI is determined for one or more Appendix I constituents at a compliance well, 40 CFR § 258.54(c) sets out what happens next:
- Within 14 days, the owner or operator places a notice in the facility's operating record saying which constituents showed statistically significant changes from background, and notifies the state that the notice is there.
- Within 90 days, the owner or operator establishes an assessment monitoring program. In assessment monitoring the wells are sampled for the much longer list of constituents in Appendix II, and results are eventually compared with groundwater protection standards rather than only with background.
- Alternatively, the owner or operator may show that the SSI was not caused by the landfill. This is often called an alternative source demonstration.
Approved states may set different schedules for these steps, so the 14 and 90 days are the federal defaults.
An SSI does not by itself mean the landfill caused it
The rule itself recognizes that an SSI can have other causes. The alternative source demonstration may show that a source other than the landfill caused the contamination, or that the SSI came from:
- an error in sampling,
- an error in laboratory analysis,
- an error in the statistical evaluation, or
- natural variation in groundwater quality.
The demonstration has to be documented in a report certified by a qualified groundwater scientist or approved by the state, and placed in the operating record. If it succeeds, the facility may continue in detection monitoring. If a successful demonstration has not been made after 90 days, the facility must begin assessment monitoring.
This is one reason the record behind an SSI matters as much as the SSI. Showing that a result came from a statistical error, or from natural variation, means being able to show exactly what was compared with what.
How Koyna handles this
Koyna computes each comparison from the stored results, and for each result it compares with a prediction limit it shows the limit, the method, and the background the limit was computed from. It flags a potential increase and shows whether the facility's retest rule calls for a resample. It does not decide that an SSI has occurred. That decision belongs to the professional who signs the report.
If something here does not match how your facilities handle an SSI, we would like to hear it. Write to support@koynasystems.com.
Sources
- 40 CFR § 258.53(g) and (h), the statistical methods allowed and their performance standards; § 258.53(i), determining whether there is a statistically significant increase over background. eCFR
- 40 CFR § 258.54, the detection monitoring program: Appendix I, semiannual sampling in (b), and what follows an SSI in (c). eCFR
- 40 CFR § 258.55, the assessment monitoring program and Appendix II. eCFR
- 40 CFR § 258.50(h), which lets approved states set alternative schedules for these steps. eCFR
- US EPA, Statistical Analysis of Groundwater Monitoring Data at RCRA Facilities: Unified Guidance, EPA 530-R-09-007, March 2009. Section 19.1 on retesting strategies. PDF