Skip to content

Chapter 4 — Planning a Tailings Monitoring Program

4.1 Planning is a process, not a checklist

A monitoring program that is pasted onto a facility after construction is weak. The strongest programs are planned risk-informed and consequence-driven from the earliest stages of design and revised as the facility evolves. Planning answers four questions before any instrument is bought:

  1. What could go wrong? (failure modes — Chapter 2)
  2. What would the early warning look like? (the measurable signal of each mode)
  3. What decisions will the data support? (trigger levels, responses)
  4. What is the right balance of visual and instrumental surveillance? (Chapter 3)

4.2 Anchor the plan to failure modes

Every instrument should trace back to a specific failure mode or a design assumption being confirmed. A pore-pressure monitor on a downstream toe exists because static liquefaction and piping are credible; a crest survey exists because slope instability and settlement are credible. If an instrument cannot be tied to a mode or a decision, question its value.

Modes first, instruments second

Resist the temptation to deploy "what we always use." Start from the facility's specific hazards — upstream construction, seismicity, weak foundation, high consequence — and let those dictate the program.

4.3 Consequence and risk categorization

Most modern tailings standards require categorizing facilities by the potential consequence of failure (often A/B/C or similar) and the likelihood given the design and controls. The categorization drives:

  • Frequency of inspection
  • Degree of automation and real-time alerting
  • Number and distribution of instruments
  • Independence and review rigor

Higher-consequence facilities warrant independent, real-time monitoring with redundant sensing, while lower-consequence facilities may rely more on competent manual surveillance.

4.4 Elements of the surveillance plan

A complete plan documents:

  • Parameters monitored (piezometric head, deformation, seepage, pond level, beach geometry, rainfall).
  • Instrument type, location, and depth with rationale.
  • Baseline and frequency of readings.
  • Trigger, action, and alarm levels (Chapter 13).
  • Roles and responsibilities for reading, reviewing, and responding.
  • Data validation and archiving procedures (Chapter 11).
  • Review and update cycle, especially after raises, storms, or earthquakes.

4.5 Staged implementation

Because a TSF is built over decades, the monitoring program is staged:

  • Pre-impoundment / commissioning — baseline survey, foundation instruments, establish datum.
  • Operating life — expand with each raise; densify where signals appear.
  • Closure — maintain surveillance through and after closure until the post-closure steady state is demonstrated.

4.6 Key takeaways

  • Plan risk-informed and consequence-driven, not by habit.
  • Every instrument traces to a failure mode or a decision.
  • Consequence category sets the intensity and automation of monitoring.
  • The program is staged across the facility's full life, including closure.