Overview of logging tools

All logging methods solve one problem: they judge the rock and the fluids from a physical response. But they take the signal from different sources. Passive methods record the formation's own fields. The gamma ray log (GR) counts photons (particles of radiation) of the natural gamma radiation of potassium, uranium and thorium. The spontaneous potential method (SP) records the electrical potentials (voltages) that arise by themselves at the boundaries between the mud filtrate, the formation water and shales. Thermometry measures the temperature in the borehole. Passive methods have no source of their own. They are simple and reliable, but they measure only what nature has provided.

Active methods excite the rock themselves and record the response. Electrode sondes pass current through the rock. Induction sondes induce eddy currents (circular currents) in it. Acoustic sondes send elastic waves (sound waves). Gamma-gamma density logging irradiates the rock with gamma rays, and neutron logging irradiates it with neutrons. Nuclear magnetic resonance (NMR) logging magnetises the hydrogen nuclei in the pores. An active sonde has a source (an electrode, a coil, a transmitter, a capsule with a radionuclide, or a generator) and one or more receivers (electrodes, coils, sensors or detectors). Their relative position determines what the sonde measures and where.

A sonde reading is not the value at a single point. It is an average over the volume of rock that the signal crosses on its way from the source to the receiver. The farther the receiver is from the source, the deeper into the formation the signal path goes. So the spacing (the length of the sonde) sets its depth of investigation. But a long sonde also averages over a longer vertical interval, so it tells thin beds apart less well: its vertical resolution is coarser. This is a general rule for all methods. The topic Open-hole logging: why, how and when explains how it shows on thin beds.

Nuclear methods have a third factor: counting statistics. The farther the detector is from the source, the fewer gamma rays or neutrons reach it, and the larger the random scatter of the counts. This is why nuclear sondes are shorter than electrical ones, and why we log them more slowly. In a combined tool string the slowest method sets the logging speed, usually a nuclear or NMR method.

By the way they sit in the borehole, tools are either pad tools or mandrel tools. In a pad tool the sensors sit in a pad that an arm presses against the borehole wall. A minimum of mud then remains between the sensors and the rock. Methods with a shallow depth of investigation need this: density logging, microresistivity tools and borehole imaging tools. The arm doubles as a caliper. The price is sensitivity to mudcake and to irregularities of the wall, and loss of contact in washouts. In a mandrel tool the sensors are inside the body. We either centralise the body (keep it in the middle of the hole) or hold it off the wall at a small standoff, as for the sonic and induction sondes. Or we press it against the wall with a spring, as for the neutron tool. We then allow for the effect of the mud with corrections.

To eliminate the effect of the borehole, we fit a sonde with two detectors or receivers at different distances from the source. Such a sonde is called compensated. The near detector is more sensitive to the mudcake, the mud and wall irregularities. The far one is more sensitive to the formation. Comparing their readings removes the disturbance that both share. The density tool finds the mudcake correction from the pair of counts. The neutron tool takes the ratio of the counts of its two detectors. The compensated sonic sonde averages the measurements of two symmetrical transmitter–receiver arrangements, and so eliminates the effect of tool tilt and of changes in borehole diameter.

No single method answers the question alone. We determine lithology and porosity from several curves with different physics. We find the fluid type by comparing porosity with resistivity. We check the suspicious readings of one method with another. That is why the basic suite for a productive interval almost always includes the GR, the caliper, resistivity of different depths of investigation, and the density–neutron pair. In English-language practice, such a set recorded in one tool string is called the “triple combo”. With the sonic log added, it is the “quad combo”.

Rock-forming minerals and characteristic rocks give recognisable combinations of readings. For the minerals, the table gives the values for a dense rock without pores. Pores filled with water reduce the density, slightly lower the photoelectric factor (a number that depends mainly on the minerals in the rock) and increase the interval transit time (how long sound takes to cross one metre of rock). The more pores there are, the stronger these effects (for more, see the section on the mixing of responses below).

We read such combinations as a whole. A low GR says only that the rock is not shaly: a clean sandstone, a limestone, anhydrite and salt all look like that. A density of 2.98 g/cm³ with $P_e$ of about 5 and an interval transit time of about 164 µs/m is anhydrite. A density of about 2.03 with $P_e$ of 4.7 and 220 µs/m is rock salt. A very low density with a high interval transit time and a low GR is coal.

Most porosity methods obey a simple rule. The reading of a sonde is made up of the readings of the components of the rock: the matrix minerals, shale and the fluids in the pores. Each reading is weighted by the share of that component in the volume the sonde sees. The step from curves to numbers for porosity and rock composition is built on this rule.

For example, a sandstone with a porosity of 25%, whose pores in the flushed zone are filled with fresh filtrate, gives a bulk density of about 2.24 g/cm³ and an interval transit time of about 292 µs/m. A limestone of the same porosity gives about 2.28 g/cm³ and 272 µs/m. These differences are small compared with the effect of porosity itself. So we cannot tell a sandstone from a limestone by density alone or by interval transit time alone. The photoelectric factor, however, differs between them almost threefold: about 1.6 and 4.5 barn/electron.

The density and neutron sondes work with sealed radioactive sources: hermetic capsules that contain a radionuclide. The gamma-ray source of the density tool is caesium-137. Its photon energy is 662 keV, and its half-life (the time in which the activity falls by half) is about 30 years. The neutron source is a mixture of americium-241 and beryllium. The alpha particles of americium knock fast neutrons with an average energy of about 4 MeV out of the beryllium. The half-life of americium-241 is about 430 years. The activity of logging sources is tens of gigabecquerels for the caesium source and hundreds of gigabecquerels for the neutron source (of the order of 1.5–2 and 16–20 curies).

A capsule source cannot be switched off: it always emits. So at the surface it is kept in a shielded container and handled with a special tool. Increasingly, a pulsed neutron generator replaces the capsule. This is a compact accelerator that emits only while voltage is applied to it. Methods for monitoring saturation in cased wells are based on such generators. Some modern tools, including those for logging while drilling, dispense with capsule sources altogether.