Five Tricone Bearing Failure Patterns: How Downhole Evidence Points to Operating or Selection Errors

A pulled tricone bit tells you why it stopped. Tight cones, sand-polished seal faces, axial play above the new-bit reading, or bearing balls on the shaker screen - each one is a different failure mechanism, and each one points back to a decision made on the rig floor days earlier.

How Bearing Type Sets the Hour Budget

Three bearing families and the hours each one buys

Roller cone bearings fall into three families, and the gap between them is structural rather than cosmetic. An open bearing lets well fluid reach the rolling elements, so the ball set and roller set run in whatever the circulation delivers. A sealed roller bearing closes the raceways behind an elastomer seal and a grease reservoir that is pressure compensated against the annulus. A sealed journal bearing replaces part of the rolling element train with a journal bushing, a thrust washer, and a hard-faced thrust face, then seals the whole assembly. In clean water an open bearing runs 80-100 hours; in solids-laden fluid the same hardware drops to 40-60 hours because angular sand cuts the raceways. Sealed families hold their numbers far better at 100-150 hours for sealed roller and 130-180 hours for sealed journal in the same fluid. That is the 2.5-3 times gap that makes a sealed bearing the default choice once a hole goes past 120 m.

What the deeper hole does to the remaining margin

The third digit of the IADC code records the bearing family: 1, 2, and 3 are non-sealed, 4 and 5 are sealed roller, and 6 and 7 are sealed journal. Water-well fleets buy most bits in 517, 537, and 617 for soft-to-medium formations, where the bearing specification and the cutting structure are matched to the same job. A 120 m hole puts 10-15 pure drilling hours on one bit, so even an open bearing spends under 25 percent of its hour budget and nobody notices the difference. Push the same hole to 200-260 m and that bit sees 22-35 hours spread across two or three shifts, while a round trip to change it costs another 1-2 hours of hoisting and connections. Bearing margin stops being a line in a catalogue and becomes the factor that decides whether the bit is retired at the right hour.

Table 1. Hours are field averages for 7-7/8 in to 9-1/2 in (200-241 mm) bits in water-base circulation, drawn from customer post-run reports. The sealed families carry roughly 2.5-3 times the open-bearing life wherever the fluid carries solids.

Seal Wash from Solids-Laden Circulation

Pattern one and why solids control comes first

Pattern one is seal wash, and it starts outside the bit. Water-well rigs rarely carry the solids-control train a mining rig would. Most run two settling pits and perhaps a 100-120 mesh shaker, which leaves sand content at 1-2 percent by volume at the bit face. Abrasive-laden fluid reaches the cone-to-leg gap at nozzle exit velocity, cuts across the seal face, and once the seal loses its line of contact the bearing is running in drilling fluid rather than grease. Field evidence is specific. The seal face reads bright and polished with radial scoring, heaviest on the side facing the jet. Where the design lets you see it, the grease comes out grey-black and emulsified. One cone turns noticeably rougher than the other two, and column five of the IADC dull sheet - the crew habit calls it column B - climbs from 2 to 6 inside 30-40 percent of rated hours. The correction sits upstream of the bit: a desander handling the 40-74 micrometre fraction together with a desilter covering 15-40 micrometre brings sand content under 0.5 percent by volume, and a sand-content test takes two minutes on the rig. Where operators cut sand from 2 percent to around 0.4 percent, customer post-run reports show hours returning to the 100-150 band instead of the 45-70 band that follows unchecked solids.

Reading the numbers that already sit on the rig

None of this needs new hardware before it starts returning hours. The sand-content kit already hangs in the doghouse on most crews, and reading it once per shift against the pump rate tells you whether the pits or the removal equipment is losing ground. Circulation volume frames the choice of removal gear: a hole circulating at 600-1200 L/min needs desander cones sized to that throughput, not rated for a smaller pump, and a cone running outside its design flow separates poorly at both ends of the particle range. Settling pits still carry weight when they are managed rather than assumed, and a pit volume that has been allowed to drop below 70 percent of design loses its residence time and puts the coarse fraction back down the hole. Crews who log sand content next to rotating hours find the correlation within ten bits, and after that the argument about whether seals or bad luck cause bearing loss ends quickly.

Cone Lock and Cone Loss on the Same Bit

Pattern two grinding instead of crushing

Pattern two is cone lock. The bit comes off bottom with one cone that will not turn by hand while the other two spin, and the inserts read evenly rounded rather than chipped. That pair of observations is the signature of grinding rather than crushing. Run an 8-1/2 in (215.9 mm) bit at 120-150 rpm on 18-22 kN and the tungsten carbide inserts skate across the rock surface, generating contact heat and polishing the formation instead of fracturing it, while the bearing carries full rated speed with almost no load to justify it. The correction is arithmetic a driller can apply in one stand: pull rotary speed back by 25 percent, taking 130 rpm to 95-100 rpm, and raise weight on bit by 10-15 percent, taking 20 kN to 22-23 kN on the same size. Customer post-run reports show the locked-cone entry disappearing once speed and weight are paired this way, because depth of cut per revolution moves out of the sub-0.5 mm polish zone and into the 1.0-1.5 mm band where inserts actually fracture rock.

Pattern three when the hour budget runs out

Pattern three is cone loss, and it is the expensive one. A locked cone costs a bit; a lost cone costs a bit plus a fishing job. As the bearing reaches the end of its life the journal wears oval, the retaining balls lose their seat, and the cone walks off the leg and drops to bottom. Surface evidence is unmistakable: one cone missing, its leg journal polished with a distinct step, and 2-4 kg of ball bearings, rollers, and locking steel spread across the shaker screens. On an 8-1/2 in bit roughly a third of the cutting structure stays in the hole, so any further rotation drills steel rather than rock. The IADC dull sheet records it honestly - column three takes LC for lost cone or BC for broken cone, and the bearing column reads 8, meaning all of the life budget went out with it. Crews report 2-5 days lost to junk basket, magnet, and spear runs, at 1,800-2,600 USD per rig day on typical water-well equipment.

Seal Lip Selection and Leg Fatigue Cracking

Pattern four the specification rather than the circulation

Pattern four is not the circulation; it is the seal specification bought for the circulation. Two constructions dominate the sealed families. A single elastomer lip seals radially against the cone bore and serves well where sand stays below 0.5 percent and rotating hours stay moderate. A dual-lip construction backed by a metal-to-metal face seal puts a hard sealing face ahead of the elastomer, so angular sand meets metal first and does not reach the rubber until the outer barrier is already past saving. Compound choice runs alongside geometry: nitrile rubber handles continuous service to roughly 120 degrees Celsius and hydrogenated nitrile models stretch that toward 150 degrees Celsius, while neither survives long once abrasive reaches the lip. Where rigs hold sustained circulation above 1 percent sand, customer post-run reports put the dual-lip plus metal face combination at 1.5-2 times the life of a single lip in the same hole. Field evidence for this pattern is asymmetric wear, where one side of the seal face reads bright while the opposite face looks barely used, meaning axial play let the lip cock under load.

Pattern five vibration-driven fatigue across interbedded rock

Pattern five is leg break, and it is driven by vibration rather than by hours. Interbedded sequences - sandstone over shale over sandy shale inside a single metre - load each cone differently on every revolution. Crews running triaxial vibration subs record axial acceleration peaks of 5-10 g through these sections against 1-3 g in a homogeneous interval, with weight on bit swinging plus or minus 50 percent as the bit bounces off each hardness change. The evidence sits on the fracture face: beach marks radiating from the crack origin, with part of the surface oxidised and dark where the crack grew slowly across many hours before the final tear. That dark region is proof that the leg failed by fatigue rather than by one overload event. Section-specific operating is what changes it. Run 60-80 rpm through the interbedded interval instead of 100-130 rpm, hold weight steady enough that the gauge swings less than 15 percent, and put a shock sub or an elastomer damping element in the bottom hole assembly so the peaks flatten before they reach the leg.

Table 2. All five patterns reward the same habit of mind, which is reading the bit on the boards before someone hoses the evidence off.

Field Checks, Prevention Ledger, and Fishing Readiness

Three checks that take eight minutes on the boards

Three checks at the bit breaker separate a worn bit from a failing bearing, and none of them needs more than a dial indicator and a magnet. Check one is rotation. Gloves on, turn each cone through a full revolution and note drag plus any hard spot; a cone that will not turn, or one that grinds at the same place every time, has a bearing problem no matter how good its inserts look. Check two is axial play. Measure cone end movement against the new-bit reading printed on the shipping tag, where new values typically sit at 0.15-0.50 mm and most operators retire or re-evaluate above 1.5 mm. Check three is the magnet. Sweep the shaker returns and separate the strongly magnetic bearing steel from the weakly magnetic carbide fragments, since cobalt-bonded carbide pulls weakly where a ball bearing snaps down hard. Logging all three against the bit serial number takes eight minutes and converts recollection into a trend.

Prevention ledger and what fishing actually costs

Prevention runs on three things that cost very little. First, buying sealed rather than open, because the 2.5-3 times hours advantage pays for itself on the first bit that goes past 15 drilling hours. Second, a life ledger with bit serial, formation, rotating hours, depth in and out, and the bearing column reading on one line, which tells a superintendent after twenty bits exactly where each IADC code runs out on his own ground; retirement at bearing column 6-7 rather than 8 keeps the cones on the bit. Third, fishing readiness, meaning a junk basket, a magnet, and a spear staged at the rig instead of trucked in later. When a cone does come off, crews with tools already on site have retrieved it in half to one day, while crews waiting on a service company report 2-5 days lost at 1,800-2,600 USD per rig day. We build sealed roller and sealed journal lines from 5-7/8 in to 12-1/4 in, match the seal construction to the sand load the rig actually carries, and ship in 15-20 days.

Table 3. Every line is something the rig already owns or already measures; none of them calls for new iron.

Bearing failure is not random. A bit that comes off a 200 m hole with one tight cone, bright seal faces, and 1.8 mm of axial play is reporting exactly what it was fed, whether that was dirty circulation, too much rotary speed for the weight carried, or too many hours for the bearing family bought. Read the three checks against the serial number in the same eight minutes, log the IADC bearing column honestly, and by the fifth bit the ledger tells you what the first one never could. That is how a parts bin stops absorbing surprises and how the next bit order carries the changes the previous five bits have already paid to discover.

Sealed tricone bearings buy 2.5-3 times the hours of open bearings in sandy fluid. Solids control, 25 percent less speed with 10-15 percent more weight, reading the IADC bearing column honestly, and fishing tools staged at the rig remove all five failure patterns.

E-mail :

yanceychang@zzsungood.com

Phone/Wechat：

+86-15037166264