Tubing And Casing Coupling

Per API 5CT Table 22: BTC coupling OD tolerance = +3%/‑1%. For 7" BTC (nominal OD 7.656"), the acceptable range is 7.579" to 7.886". Coupling length tolerance = ±1/8" (nominal 10" for BTC). Wall at pin nose: 0.375" minimum.

30-second dock check: gauge OD at both ends with a snap gauge set at 7.579" (low) and 7.886" (high). If too tight to pass the low gauge, or too loose on the high, reject it. Tape the length — 10-1/8" is fine; 10-1/4" means the facing pass was skipped and the coupling won't make torque correctly. Hardness spot-check per Table E.2: P110 max 32 HRC, N80-Q max 25 HRC. Three numbers, thirty seconds, one red flag.

Round thread (STC, LTC): the V-profile leaves a controlled 0.002–0.005" clearance between crest and root along the entire helical path. Gas follows that spiral gap to the surface — seal depends entirely on thread compound (API Bul 5A2, 60–75% Zn dust) plugging the leak path. No compound = guaranteed leak above 500 psi.

BTC (Buttress): the thread has a negative flank angle (3° stab flank, 10° load flank) with a flat crest that contacts a flat root when made up. When torqued, the flanks compress and the crest contacts the root — creating a metal-to-metal seal. BTC holds gas at 3,000–5,000 psi without relying on compound. Premium connections go further: they add a separate torque shoulder and a radial metal-to-metal seal that operates independently of the threads, rated gas-tight at ≥95% of pipe body pressure per ISO 13679 CAL IV. You pay for what you need — compound-dependent, metal-to-metal, or independent seal.

API 5CT Table E.2 coupling hardness limits per grade: J55 = not required (as-rolled). N80-1 = not required. N80-Q = max 25 HRC. L80-1 = max 23 HRC with temper ≥621°C. C95 = max 26 HRC. T95 = max 26 HRC plus SSC testing per NACE TM0177. P110 = max 32 HRC. Q125 = max 35 HRC (subgrade-dependent).

The detail most people miss: coupling hardness is measured on the box face — three readings per end, averaged. Coupling body hardness typically runs 2–4 HRC higher than the pipe body because the thicker coupling wall retains heat longer during Q&T. A P110 coupling reading 35 HRC on the box face passes API but fails NACE MR0175. Always check where on the MTC the hardness was taken. If it only says "HRC 32" without a location, ask.

API 5CT Section 6.2 plus API Spec 7B-11 say: BTC taper = 0.75 in/ft (0.0625 in/in) with tolerance +0.003/‑0.000. Pitch diameter tolerance = ±0.002" per Table 23. Thread height tolerance = +0.000/‑0.002". In theory, two API-compliant parts from different mills should make up. In the real world, they often don't make up well.

The problem: manufacturing bias. Mill A's pins consistently land at +0.001" on taper; Mill B's boxes at −0.001". The resulting 0.002" gap produces a loose fit that loses 20–30% of seal contact stress. Safe rule: use matched suppliers for new strings. For replacement single joints, always test-make-up one sample before running 100 into the hole. If the torque-versus-turns curve is abnormal — especially a flat plateau or late shoulder — compared to the rest of the string, pull it. One bad connection in a 500-joint string costs far more than the test make-up.

Yes. Per API RP 5C5 Table A.3, BTC 7" 26 lb/ft make-up torque = 3,450 ft-lb at 20°C. Steel's coefficient of thermal expansion is 11.7 × 10⁻⁶/°C. At a 60°C temperature rise (20°C to 80°C BHST), the axial thermal load on the BTC connection = E × α × ΔT × cross-sectional area = 30 × 10⁶ × 11.7 × 10⁻⁶ × 60 × π/4 × (7.656² − 6.276²) = roughly 28,000 lbs of additional tensile load on the connection.

That 28,000 lbs acts against the compressive preload the make-up torque created — it effectively reduces the effective seal preload. For wells with BHST above 80°C, increase make-up torque by 5–10%: 3,450 × 1.05 = 3,620 ft-lb. For premium connections with separate metal-to-metal seals, follow the manufacturer's temperature-adjusted torque tables. Most call for 3–7% more per 50°C above ambient. Don't guess — calculate, or you risk a gas leak at temperature that was pressure-tested cold.

API 5CT Section 14 defines reuse criteria. Run through them in order:

(1) Visual: no cracks, cuts, or corrosion pitting >0.1 mm deep per Table E.4.

(2) Thread gauge: accept if taper and pitch diameter stay within +0.005" of new spec, thread height loss ≤25% per API RP 5C5 §4.

(3) OD: minimum per Table 22 — 7" BTC = 7.579".

(4) Remaining wall at pin nose: ≥90% of new = 0.338" minimum. Below that, crack risk at the stress-relief groove makes it scrap.

(5) Hardness re-check: L80 still ≤23 HRC, P110 ≤32 HRC per Table E.2.

In practice, a coupling from a sweet well with no corrosion history usually passes all five. The thread is always the weak point — if the gauge fails or visual shows tooth wear past 25%, rethread it. Maximum two rethreads per coupling; beyond that, the wall becomes too thin. Re-cert runs $3–5/coupling versus $25–40 new. For established strings, it's worth it. For critical wells, don't recycle — the inspection cost saving is not worth the connection gamble.