You’ve run the Surface Safety Valve (SSSV) function and leak test. The chart shows a slow bleed-off, but it’s within the industry-standard acceptance limits: maybe 380 cm³/min on a liquid well, or 14 scf/min on gas. The paperwork gets a tick, and the well is good to go. But is it?
The hard truth is that a valve just scraping by the acceptance criteria is often a valve telling you it has a problem. Relying on a marginal “pass” can give a false sense of security, setting you up for a more significant failure down the line when the well really challenges that valve.
The Engineering Reality of SSSV Leakage
A routine SSSV leak test demonstrates one thing: that on that specific day, at that differential, in that direction, with that fluid, the valve closed and held within a defined leakage allowance. It does not demonstrate the valve will close and hold during a genuine emergency, when the differential is higher, the flow is at full rate, and the fluid is carrying whatever the well is producing.
The industry acceptance allowances—typically around 400 cm³/min for liquid and 15 scf/min for gas—are widely quoted but often misunderstood. These are acceptance limits, not performance targets. They were set to distinguish a failed valve from a functioning one, with practical allowances for wellsite measurement. A healthy valve in good condition leaks far below these limits, often not measurably at all.
Consider a valve reading 380 cm³/min. While technically “within spec,” it’s leaking two orders of magnitude more than a healthy valve. This isn’t a “fine” result; it’s a warning. Furthermore, leakage through a damaged seat isn’t linear with differential pressure. A valve passing at a modest test differential can leak substantially more at the higher differentials imposed by a genuine shut-in, as the leak path opens under increased load.
The critical distinction is between compliance and engineering judgment. A simple pass/fail answers a compliance question. Trending the actual measured leak rate answers an engineering question. If a valve consistently read zero leakage for three years and now shows 120 cm³/min, it has effectively failed in the sense that matters, even if it technically “passes” the limit. Always record the measured rate, not just the pass/fail outcome.
Running the Test So the Result Means Something
Most unreliable safety valve data stems from procedural shortcuts, not instrument error. To get meaningful data, attention to detail during the test is paramount.
Stabilisation is Key
After closing the valve, the tubing above it must be bled down to the test differential and then left alone long enough for thermal and compressibility transients to settle. Bleeding down and immediately starting a timed measurement will record fluid expansion or contraction, not true leakage. On a gas well, the settling period is longer than intuition suggests because the gas column above the valve cools as it expands. Rushing this step will give you a false reading.
Test Direction Matters
The SSSV is designed to hold pressure from above, against the direction of well flow. Therefore, it must be tested from above. The test tells you nothing about the valve’s ability to hold pressure from below. This is critical when planning operations that put pressure under a closed valve—such as a bullhead, a squeeze, or an annulus-to-tubing communication test. Do not rely on the SSSV as a barrier in these situations.
Accurate Measurement Method
Volumetric bleed-back into a measured container is the most honest and reliable method for assessing leakage. Inferring leakage from a pressure build-up above the valve is convenient but conflates actual leakage with thermal expansion and any leaks past the tree valves. If you use the pressure method, a separate, mandatory check that the tree valves are tight is essential. Otherwise, a leaking swab valve can easily be recorded as a leaking SSSV.
Consider this worked example:
- Above a closed SSSV, there is 900 m of 3½ in tubing containing brine.
- Internal volume is approximately 0.0044 m³/m, so the trapped volume is roughly 3.96 m³ (3,960 litres).
- Brine compressibility is about 3.0 × 10⁻⁶ /psi.
- A pressure rise of 40 psi from a temperature transient corresponds to a volume change of: 3,960 litres × 3.0 × 10⁻⁶ /psi × 40 psi ≈ 0.475 litres.
This means 475 ml of apparent influx can be fully explained by a 40 psi thermal transient in an entirely tight valve. Over a 15-minute observation, that’s an apparent 32 cm³/min of “leakage” from nothing at all. On a liquid-filled tubing, the pressure method cannot reliably resolve leakage below roughly 100 cm³/min unless the temperature has genuinely stabilized. When the number matters, use volumetric bleed-back.
Interpreting Marginal Results and Lessons Learned
A “pass” is not always a pass. Understanding the nuances of your test results can prevent future failures.
- Zero, consistently, then a step to a measurable rate: This often indicates seat damage from a specific event—debris, a wireline run, or a solids slug. Review what happened between tests and consider inspection at the next opportunity.
- Slow rise over several tests: This suggests erosion or scale build-up on the sealing surface. Extrapolate the trend to predict when the limit will be reached and plan intervention before that point, not after.
- Passes on liquid, fails or is erratic on gas: This points to a small seat imperfection. Gas will find paths that liquid might bridge. Treat the gas result as the true condition of the valve.
- Fails to close, then closes on second attempt: This is a critical warning sign of a sticking flapper, debris, hydrate, or a slow control system bleed. Do not record this as a pass. A real event offers no second chances. Investigate the closure delay immediately.
- Closure time longer than previous tests: This indicates a developing problem like a control line restriction or spring degradation. Time the closure on every test and trend it. This is a better health indicator than leak rate alone and costs nothing to measure. Operators who do this catch failures a year earlier than those who don’t.
The standard test will not tell you everything. A routine SSSV test has several limitations:
- Closure against flow: The routine test closes the valve into a static or near-static column. Slam closure against full production rate is a different mechanical event, qualified at manufacture but never repeated in the field.
- Debris tolerance: The test uses whatever is in the well at the time, which on a shut-in well is the cleanest it ever gets. A real event might involve significant debris.
- Control system response to a genuine trip: Testing by manual bleed at the panel does not exercise the same logic path as an Emergency Shut Down (ESD) initiation. Periodically testing via a genuine trip signal exercises the whole chain.
- Long-term hold: A fifteen-minute test says little about a valve required to hold for days during an evacuation.
For a complete record, ensure you capture the following minimum field data:
- Date of test
- Test differential pressure
- Direction of test
- Fluid type
- Measured leak rate with units
- Measurement method
- Time from signal to confirmed closure
- Control pressure at closure
- Control pressure required to reopen
- Whether closure was achieved on the first attempt
- The previous three test results for comparison (print them alongside for easy reference)
The Bottom Line
The SSSV is the only barrier element in the well required to change state under emergency conditions, and the routine test exercises it under the gentlest conditions it will ever see. Treat every measurable leak and every slow closure as vital information about a valve moving toward unavailability. By doing so, the six-monthly test becomes a genuine predictive tool rather than just a periodic confirmation that nothing has failed completely yet. Have a question about your well? Reach out via the contact page.