The Economics of DTH Button Bit Reconditioning: When to Regrind, When to Replace, and the Cost per Meter Math

A button bit does not wear evenly. The gauge row loses height fastest, so by the time the face buttons show a coin-sized flat the outer row has dropped below the height where carbide still has body behind it. Regrind inside that window and the bit returns 60-80 percent more hole from steel the contractor already owns.

The Wear Flat and the Window Around It

How a button actually loses height

A 4 in DTH bit carries 8 to 14 tungsten carbide buttons pressed into a hardened alloy steel body, with the face buttons typically at 13-14 mm diameter and the gauge row at 14-16 mm. New, roughly half of each button stands proud of the body face: a 16 mm button leaves 8-9 mm of carbide exposed, and that exposed length is the only part doing work. Under a hammer running at 14-24 bar and 1,500-2,000 blows per minute, the button does not stay spherical. It flattens where it strikes rock, and the flat grows in diameter while the exposed height drops. What makes the decision harder is that the two rows do not flatten at the same rate. The gauge buttons travel the circumference of the hole and take glancing contact against the wall on every revolution while also cutting bottom, so in granite and quartzite they lose height 30-40 percent faster than the face row. A bit can look serviceable on the face and already be below the safe limit at the gauge.

The 40 to 50 percent window and what closes it

The rule contractors settle on after running a few dozen bits is simple to state and measured in exposed height rather than in appearance. Regrind when the gauge row has lost 40-50 percent of its original exposed height, which on a 16 mm button means the 8-9 mm of carbide standing proud has dropped to roughly 4-5 mm. The physics behind the number is contact area. Once the flat on a working button exceeds about 25 percent of its crown area, the same hammer energy is delivered through a larger footprint and impact energy transfer efficiency drops by 15-30 percent, which is the point where penetration rate falls faster than the bit is wearing. Crews read it in the field without instruments: penetration rate in a consistent formation drops 20-25 percent with no change in air pressure, rotation, or pulldown, and the sound at the collar changes from a sharp crack to a duller thud. Wait beyond the window and the second problem arrives, because a shortened button carries less carbide above the body face and the grinder has nothing left to take off without exposing the press-fit interface.

Table 1. Thresholds are for 4 in and 6 in bits in granite and quartzite at 150-220 MPa unconfined compressive strength. Softer formations stretch every band; abrasive sandstone compresses them.

Four Defects That End a Bit Before Grinding Does

Button root exposure and body wash

The first defect is not in the carbide at all, it is in the steel holding it. Cuttings and rock flour driven back across the bit face at hammer exhaust velocity cut into the body between the buttons, a process crews call body wash. Once the body face erodes 3-4 mm below its original plane the buttons lose the surrounding steel that carries their lateral load, and the press-fit grip falls away. The field sign is a gap you can see: run a fingernail or a scribe around the base of a button and if it catches in a groove, the steel has gone and no amount of grinding restores it. Grinding in that condition is worse than useless, because removing more carbide height puts a shorter button into an even weaker seat. The same rule applies to a button that has sunk rather than worn, where the crown reads at or below the body plane while the neighbouring buttons stand proud. A sunken button means the hole behind it has oversized, and that bit is finished.

Cracks, thread damage, and skirt fatigue

The second defect is cracking, and it hides under the cuttings film that a freshly pulled bit carries. Wipe the face clean and look at each button under a light held low and to the side: heat checking reads as a fine network of lines across the crown, while an impact crack runs single and radial from the crown toward the base. A bit with cracked buttons goes to scrap, because grinding load pushes the crack through the remaining carbide in seconds. The third defect sits at the shank. DTH bits thread into the hammer on an API or a proprietary connection, and that thread carries the full blow energy plus rotation torque every second of the run. A thread gauge tells you within a minute whether the flank has worn past 0.5 mm or developed a step at the load flank, and either one means the bit will not hold in the chuck. The fourth is skirt fatigue, where the outer steel skirt of the bit cracks radially from the gauge row upward. It shows as a hairline with rust bleeding out of it on a bit that has sat in the rack for a week, and it is the defect most often missed because it reads as dirt until the bit is cleaned.

Table 2. Six checks take under ten minutes per bit and decide every regrinding decision a contractor makes. Skipping the last three is how bad bits reach the grinder.

What Regrinding Costs and What It Returns

Cost per meter across three lives

Professional regrinding runs 15-40 USD per bit depending on size and how much carbide has to come off, against 400-700 USD for a new 4 in bit and 700-1,500 USD for a new 6 in. The arithmetic that matters is not the comparison of those two numbers but the cost per metre over the whole life of the steel. Take a 4 in bit bought at 550 USD that drills 200 m of granite before it reaches the window. Reground once for 25 USD it returns roughly 100 m, reground a second time for 25 USD it returns about 50 m. The bit has now produced 350 m from 600 USD of steel, carbide, and grinding, which is 1.74 USD per metre against the 2.75 USD per metre the same hole would have cost on new bits only. Total life has gone up 60-80 percent and the bit cost per metre has come down 25-40 percent, and both bands widen in softer rock and narrow in abrasive quartzite. The third pass is where the model stops working, because the remaining carbide above the body face is too short to survive another 2-3 mm of removal.

The time column most cost models leave out

Bit cost is not the whole account, and contractors who regrind on the bit-cost line alone get surprised. A reground button has a larger crown radius than a new one, so it delivers the same blow through a slightly wider contact and penetration rate runs 5-10 percent lower in the same formation. On the 350 m above, running a new bit throughout would take 29.2 hours of drilling at 12 m/h plus 0.9 hours of extra rod trips to cover the 1.75 bits consumed, while the regrind path takes 30.8 hours across its three passes. The difference is 0.7 hours, and at 1,800-2,600 USD per rig day that is roughly 160-230 USD, which the 350 USD saved on bit cost still covers. Now add the variable the calculation usually omits: turnaround. A local grinding shop takes 1-3 days and an in-house grinder does the bit in 4-8 hours, but either way the bit is out of the rotation, and a fleet running two 4 in bits cannot absorb a three-day gap in the middle of a job. That is the real constraint on regrinding, and it has nothing to do with carbide. A contractor with three or more bits of the same size in rotation regrinds freely, because one is always in the hole while another is at the shop. A contractor with one bit regrinds only when the rig is already standing.

Table 3. A 4 in bit in granite at 150-220 MPa. The saving is real but small, which is why turnaround time rather than bit price decides whether regrinding pays.

Button Profile, Grinding Heat, and the Limits of a Second Pass

Spherical and ballistic buttons take different grinds

Profile decides how much material is available and what shape the grinder has to leave behind. A spherical button has the most usable carbide above the body because the crown is a full hemisphere, and it tolerates a simple pass that restores a round crown at a smaller radius. A ballistic or conical button carries its penetration advantage in the tip geometry, so grinding it into a hemisphere throws away the reason it was chosen: it turns into a spherical button with less carbide and none of the penetration it had. A ballistic button has to be reground back to a tip, meaning the wheel follows the original flank profile and leaves a small flat at the apex rather than rounding the whole crown. Most grinding shops have a template for spherical buttons and improvise on ballistic ones, so the question to ask before handing a bit over is whether they have a profile fixture for that button shape. Crews who run ballistic buttons in hard granite and get them reground round report the penetration loss going from the expected 5-10 percent to 15-20 percent, which is the whole saving gone.

Heat checking and how a grinder avoids causing it

Carbide fails in a grinder for the same reason it fails downhole, and both come down to temperature. Tungsten carbide expands at about 5.2 x 10-6 per kelvin while the cobalt binder expands at roughly 13 x 10-6 per kelvin, so a button heated unevenly puts the binder phase in tension against the carbide skeleton. Push the surface past 600-700 degrees Celsius and the binder softens, the differential opens microcracks, and the button comes off the wheel already heat checked, ready to spall on the first hard blow. The controls are unglamorous and non-negotiable. Grind with a resin-bonded diamond wheel in the 100-200 mesh range rather than an aluminium oxide wheel, which generates heat instead of cutting. Keep continuous coolant on the contact so the button never rises beyond warm to the hand. Take 0.02-0.05 mm per pass instead of trying to remove 2-3 mm in one cut. Leave a 0.5 mm chamfer at the new crown edge, because a sharp ground edge concentrates impact stress and chips. A shop that grinds dry, or that grinds to a shiny finish because it looks good, is producing cracks rather than removing wear.

Running a Reground Bit Into the Dry Season

Parameter changes after a pass

A bit coming back from the grinder is not the bit that went in, and running it on the old numbers is how a reground pass gets blamed for poor performance. With a larger crown radius the button needs slightly more energy per blow to reach the same depth of cut, so add 1-2 bar of air pressure where the compressor has it, taking a 4 in hammer from 18 bar to 20 bar. Hold pulldown 5-10 percent lower on the first 20-30 m while the new crown beddes in, because a freshly ground edge is the most vulnerable it will ever be. Rotation setting stays where it was; the button spacing has not changed. Watch the hole gauge rather than the penetration rate for the first hour, because a reground bit whose gauge row was ground unevenly will leave the hole undersized and the next bit will have to ream its way down. Where crews log penetration rate per bit serial, the regrind entry carries a note of how many passes the bit has had, and the second pass always reads lower than the first.

Storage rules between passes and the pre-season window

Storage between passes decides whether the second pass happens at all. A reground bit laid on bare concrete in a coastal yard picks up surface rust on the ground crown within days, and rust pits become crack origins on the next run. The rules cost nothing: thread protector caps on every bit in the rack, a thin film of rust-preventive oil on the ground face and the shank, and stack no more than three high on a timber bearer rather than directly on the floor. Carbide does not rust but the steel around it does, and corrosion at the button base is what turns a serviceable bit into scrap between seasons. The timing question sits on the same logic. Across most of sub-Saharan Africa the dry season runs November to March with the construction peak in December to February, which leaves October as the last month when rigs are still working at reduced load and roads are still open. A bit sent to the grinder in October comes back into a rack that will be emptied in six weeks, with no turnaround cost because the rig was never waiting on it. The same bit sent in January costs three days of a 2,000 USD rig day and arrives into the middle of the peak. Contractors who treat regrinding as a pre-season task rather than a mid-job repair are the ones who get the second and third pass for free.

Table 4. The decision is a function of fleet size and calendar position more than of bit price, which is why two contractors running the same rock reach opposite answers.

The honest summary is that regrinding saves less money than the brochures suggest and more downtime than anyone budgets for. Bit cost falls 25-40 percent per metre, which on a 350 m hole is about 350 USD, and rig time gives back 160-230 USD of it, leaving a net saving that would not justify the trouble on its own. What justifies it is the rack. A contractor who regrinds every qualifying bit in October enters the dry season with three or four serviceable bits where he had one, and when the next bit takes 30-45 days to arrive he keeps drilling instead of explaining the delay. We build 4 in to 8 in button bits on new bodies with fresh carbide pressed into every seat, and we ship in 15-20 days. Everything described above is what a contractor does with a bit he already owns and has already run, which is the only regrinding any bit should ever see.

Regrind a DTH button bit when the gauge row has lost 40-50 percent of its exposed height, retire it below 40 percent or at the first sign of body wash, cracked buttons, thread wear past 0.5 mm, or a cracked skirt. Two passes turn 200 m into 350 m and bring bit cost per metre down 25-40 percent. The saving is small; the reason to do it is the rack going into the dry season.

E-mail :

yanceychang@zzsungood.com

Phone/Wechat：

+86-15037166264