Where magnets appear downhole
| Application | Magnet role | Dominant constraint |
|---|---|---|
| MWD / LWD instrumentation | Magnetometer references, orientation and toolface sensing | Thermal stability of the reference field — drift becomes survey error |
| Mud pulse telemetry | Actuator and valve drive motors | Temperature plus cyclic duty in abrasive fluid |
| Electric submersible pumps | PM motor rotors in permanent magnet ESP designs | Sustained high temperature at depth, long service life, no access |
| Downhole generators / turbines | PM alternator rotors powering tool electronics | Temperature and vibration together |
| Rotary steerable systems | Actuator motors, position feedback, couplings | Shock and vibration during drilling |
| Magnetic couplings | Torque transfer through a sealed pressure barrier | Eliminating a dynamic seal at pressure — see magnetic couplings |
| Fishing and retrieval tools | Recovering ferrous debris and lost components | Holding force through mud, with reliable release at surface |
| Casing collar locators, flow meters | Sensing references | Long-term stability without recalibration |
The economics that shape every decision
A magnet is a trivial fraction of the cost of a downhole tool and a rounding error against the cost of a trip to replace one. Intervention on a deep well runs into six figures before any equipment cost, and rig time is charged by the day. This inverts the usual cost logic: specifying up is almost always correct here. The premium for samarium cobalt over neodymium, or for a higher temperature class than the nominal case requires, is insurance bought at a discount.
Temperature: the governing constraint
Geothermal gradient alone puts a deep well well beyond standard magnet territory, and circulating fluid, drilling friction and tool self-heating add to it. HPHT classifications commonly place tools in the 150–200 °C band with ultra-HPHT service extending beyond.
| Material | Practical ceiling | Br drift per °C | Downhole verdict |
|---|---|---|---|
| NdFeB, N class | ~80 °C | −0.12% | Surface equipment only |
| NdFeB, SH / UH | 150–180 °C | −0.12% | Shallow and moderate wells, with margin checked |
| NdFeB, EH / AH | 200–230 °C | −0.12% | Upper limit of NdFeB; heavy rare earth cost and licensing exposure |
| SmCo 1:5 | ~250 °C | −0.04% | Good, and inherently corrosion resistant |
| SmCo 2:17 | ~350 °C | −0.03% | The downhole default |
| Alnico | 450–550 °C | Very low | Extreme temperature, but low coercivity limits it to specific circuits |
Two reasons SmCo wins downhole, and only one is the temperature rating
The obvious one is headroom — Sm₂Co₁₇ operates where NdFeB has no usable margin left.
The less obvious and often more important one is drift. SmCo’s remanence temperature coefficient is roughly a quarter of NdFeB’s. Over a 150 °C excursion, NdFeB loses on the order of 18% of its remanence reversibly while SmCo loses around 4–5%. For a magnetometer reference or a survey instrument, that difference is the measurement error, and it appears as wellbore position uncertainty rather than as a component fault. Where the magnet is part of a measurement, stability matters more than strength.
Specify against the curve, not the datasheet
- Request demagnetization curves at operating temperature. Room-temperature datasheet values are close to meaningless for a part that will live at 175 °C. You need to see where the knee sits in your operating region, at that temperature, for that grade.
- Run the load-line analysis hot, with the actual permeance coefficient of your geometry. A thin magnet in a large air gap is far more vulnerable than the same material in a closed circuit — the method is in the temperature guide.
- Add margin for the excursion, not the average. Circulation stops, the tool sits, and bottom-hole temperature climbs. Design to the soak, not the steady state.
- Consider thermal stabilisation. Pre-baking magnetized parts above worst-case service temperature takes the one-time irreversible knock-down before calibration rather than after it — essential for instruments and worth specifying on the drawing.
- Ageing matters at these temperatures. Long-term flux loss over thousands of hours hot is a different question from short-term thermal loss. For multi-year installations such as ESP motors, ask about long-term stability data, not just the temperature rating.
Pressure, chemistry and sour service
Pressure
Hydrostatic pressure does not meaningfully affect magnetic properties — sintered magnet material is fully dense and near-incompressible. The pressure problem is mechanical and structural:
- Magnets are usually inside a pressure barrier, so the barrier and its seals are the design problem, not the magnet.
- Where a magnet sits in a pressure-balanced or oil-filled compartment, any voids or porosity in an encapsulation become collapse points.
- Differential pressure across an encapsulation can crack brittle material. Encapsulation should be void-free and pressure-compensated where the design allows.
Chemistry
Coating is not a barrier at depth
Nickel-copper-nickel plating is a corrosion coating for ordinary industrial life, not a pressure barrier or a chemical containment for downhole fluids. Anything that reaches a coated NdFeB magnet in a well will get past the plating. The design has to isolate the magnet with a real barrier — a welded housing, a qualified encapsulation, a sealed can — and the magnet coating is then a secondary defence for the case where that barrier is compromised. SmCo’s ability to survive without a coating at all is one of the reasons it is chosen.
Shock, vibration and retention
Drilling produces one of the harshest mechanical environments any electromechanical assembly faces — sustained lateral and axial vibration, stick-slip torsional oscillation, and shock from bit bounce and formation changes, all for hundreds of hours at temperature.
The material problem
Both NdFeB and SmCo are brittle, and SmCo is the more brittle of the two. It chips and cracks more readily, and the mechanical environment downhole is the worst place for that. The structural risk is not demagnetization from shock — sintered rare earth material is largely insensitive to that — it is fracture. Design accordingly.
Retention principles that apply here
- Load in compression, never in tension or bending. Compressive strength is high; tensile strength is a small fraction of it.
- No point loads. Conformal support across a face, not a clamp on a corner. Brittle material has no yielding to redistribute a stress concentration.
- Do not rely on adhesive alone. Adhesive at 175 °C under a hundred hours of vibration is a different material from adhesive on a bench. Geometry should retain the part if the bond fails completely.
- Eliminate free movement. Any clearance becomes an impact, and repeated impact on a brittle edge produces fragments — which then migrate onto every ferrous surface in the tool.
- Account for differential expansion. Sintered magnet material has anisotropic thermal expansion that matches neither steel nor the housing alloy. Over a 150 °C swing that mismatch is significant, and a joint that was sound cold can clamp or go slack hot. Check the fit at both extremes.
- Segment large magnets. Building a long rotor from stacked shorter segments reduces the consequence of any single fracture and eases thermal strain.
The full mechanical treatment, including pocket design and containment for rotating assemblies, is in assembly and retention design.
What we supply
A note on scope
Radial Magnets supplies magnet components and assemblies. Qualification of a finished downhole tool — pressure testing, sour service qualification, HPHT rating — belongs to the tool builder. What we contribute is the right material, characterised at your operating temperature rather than at 20 °C, with documentation that survives an audit. Our magnets are not sold or approved for military use.
- Samarium cobalt in both alloy systems — SmCo₅ and Sm₂Co₁₇ — in discs, rings, blocks, arcs and custom geometries.
- High-coercivity NdFeB in SH through AH classes where the thermal envelope genuinely allows it and cost or field strength favours it.
- Alnico for the extreme-temperature, low-coercivity-tolerant circuits where it still has no substitute.
- Elevated-temperature demagnetization curves for the specified grade, so the design can be checked where it will actually operate.
- Thermal stabilisation to a specified temperature before delivery, for instrument-grade parts requiring calibration stability.
- Per-lot material certification with measured properties, and full lot traceability.
What to send us
- Maximum bottom-hole temperature and the soak duration — and whether that is the circulating or the static figure
- Whether the magnet is isolated behind a pressure barrier, or exposed to well fluid
- Fluid chemistry, including H₂S, CO₂ and hydrogen exposure
- Shock and vibration specification for the tool
- Whether the magnet serves a measurement function, in which case state the allowable drift rather than the field
- Expected service life and the accessibility for replacement
- Geometry envelope and the retention method planned
Grades carrying dysprosium and terbium — and samarium cobalt in its entirety — fall under export licensing, which adds materially to lead time. On downhole programmes with long qualification cycles this is worth planning around early; see lead times and MOQs.