Wednesday 16 October 2019

Downhole logging tools: the resolution question

As with many logging-data end users, I have always been aware of the variable spatial resolutions of logging tools. I am also aware of the swathe of variables that can influence the resolution of logging data during the data collection phase. These include the source-sensor arrangement on the tool string, sampling rate, logging speed, penetrative strength of any radioactive sources, borehole conditions such as hole diameter and mud density, and the appropriateness of the tools for investigating the specific downhole lithologies. However, for the next few minutes I would like to focus solely on the tools themselves and what exactly can be measured when all other conditions are ideal.

Logging tool technology utilizes many forms of investigative method – from acoustic pulses to magnetic fields, passing through a range of different nuclear source types on the way. By consequence, individual tools measure different areas of the downhole environment. In other words, they exhibit different spatial resolutions. 

Spatial (or measurement) resolution in logging tools can be separated into vertical resolution and depth of investigation; with vertical resolution detailing the ability of a tool to resolve changes in tool response parallel to the tool axis (Schlumberger); and depth of investigation detailing tool response perpendicular to tool axis. In other words, depth of investigation tells us how far into a geological formation (i.e. beyond the borehole walls) a tools response can be influenced.

The vertical resolution and depth of investigation of any given tool are often linked, as most physical fields are spherical (Clark et al. 1988). However, this is not always the case. The sonic tool, for example, produces acoustic waves that do not propagate spherically but rather with a “field” shape closer to that of an ellipsoid (Rider and Kennedy, 2011 – strongly recommended reading for a summary of acoustic wave propagation characteristics within a formation). A good rule of thumb in modern sonic tools is that the depth of investigation in inches, is the same as the transmitter receiver spacing in feet.

The vertical depth of any given logging tool is often controlled by, or at least heavily influenced by, its transmitter-receiver spacing. However, this is only part of the story. While a tool may have excellent qualitative spatial/vertical resolution, it may still exhibit poor quantitative resolution. Quantitative tool resolution can also be termed minimum bed thickness; defined as the minimum thickness of geological bed for which a tool can/will detect true values.

Most minimum bed thicknesses are calculated through tool testing; as the minimum thickness often relies as much on the formation being tested as it does the tool itself. This creates a situation where certain lithologies can be read for true values at finer resolutions than other lithologies.

For a good example of how vertical resolution of a tool can be lithology-dependant take a look at electrical resistivity tools. Resistivity tools can have extensive vertical resolutions, but can also be subject to shoulder bed effects. This is where the beds either side of a bed of interest influence the tools response. The influence of a shoulder bed is magnified with increasing contrast between the measurement values of the two beds. Furthermore, studies by Clarke et al. (1988) show that shoulder bed effects can be greater if the shouldering beds are of lower resistivity (more electrically conductive) relative to the bed of interest. 

Altered from Clarke et al. 1988

In addition to vertical resolution, measurement depth of investigation can also be lithology-dependant. A good example here comes from the porosity tool where non-porous beds can be seen to a greater depth than a porous bed as water contained in pores intercepts neutrons emitted from the source. This effect can be significant, with 0% porosity beds penetrated by 60 cm, 10% porosity beds by 34 cm, and 30% porosity beds by only 16.5 cm (Rider and Kennedy, 2011).

There are few cases where a logging tools minimum bed thickness is in fact finer than its vertical resolution. Theys (1999) showed that the photoelectric factor output of the density tool often accurately identified beds 10-20 cm thick, even though the transmitter-receiver spacing for the tool was 36 cm and the minimum bed thickness for the overall density measurement was closer to 40 cm.

Due to this lithology control (in addition to the environmental controls omitted in this blog) it can prove challenging to report on the true vertical resolution of a given logging tool. See the image below for some rough areas of investigation of common logging tools. 

Approximate to-scale graphical representation of common logging tool areas of investigation 

In order to find out what is truly going on down a borehole, a data user needs to understand both what is being measured, and how – more reason to interrogate multiple data types – core, seismic, and log data to name a few.

Failing the ability to do this, one can always use certain log data to quality control other data since most tools utilise different investigative technologies and thus present different limitations. There are also inversion methods and other tools that data users can utilize to increase tool accuracy and shrink minimum bed thickness detection windows. A good summary of some methods can be found in this paper by Sanchez-Ramirez et al. (2009).

As I see it, the only way round this issue is (as with any trades person) to know your tools as there is nearly always a difference between bed detectability and minimum bed thickness.

Laurence


References:






Friday 19 July 2019

Sourcing petrophysics data

In my previous blog I summarised a variety of ways in which petrophysical data can provide insight into a range of geoscience fields. But what use is this if you don’t have any data to work with? So I thought this might make a useful next topic – sourcing petrophysics data.

The short list below details a few ways of acquiring publicly available petrophysics data. As I work within the IODP my knowledge-base is a little skewed, that said however, the list is in no particular order.

1. IODP

Let’s start with the obvious, or at least the most obvious to me - the International Ocean Discovery Program (IODP). The IODP is an international marine research collaboration that explores Earth's history and dynamics using ocean-going research platforms to recover data recorded in seafloor sediments and rocks, and to monitor subseafloor environments (source). Throughout its 50 year life-cycle the IODP has drilled, logged and collected samples from almost every geological setting around the world’s oceans (Goldberg, 1997). Such a useful resource that IODP data is even fed into the Neftex Earth Model, and IODP microfossil data into Nannotax.

This breadth however sometimes means that finding targeted data can be challenging for researchers that are not fluent in IODP language and structure. For this reason I will try and simplify the various databases below.

The IODP consists of three operating platforms, the JOIDES Resolution, the Chikyu and the Mission-Specific Platform (MSP) (see below). These operating platforms each have individual online databases for the data they collect. Data types stored in these databases include all data from core analyses both on whole round recovered cores (e.g. multi-sensor track system data) and split/slabbed core (e.g. core images, colour reflectance data); as well as all sample-based analyses (e.g. discrete measurements, chemical analyses etc.). For a complete list of standard measurements collected on every expedition click here.

Logging data (as opposed to core/sample-based data) recovered by IODP on the other hand is simple data to find - with all three operating platforms depositing data in the IODP log database.


Databases associated with each IODP platform
Links:
Downhole logging database
HYPERLINK "http://brg.ldeo.columbia.edu/logdb/" http://brg.ldeo.columbia.edu/logdb/scientific_ocean_drilling/ HYPERLINK "http://brg.ldeo.columbia.edu/logdb/"
MSP/Pangea
Chikyu
JOIDES Resolution:
Overview

JANUS (Pre Exp. 317 - 2009)

LIMS (Post Exp. 317 - 2009)









But what if you wish to find data on a specific lithology or structure? Or what about from a particular ocean/region? The IODP has undertaken nearly 300 scientific exploration expeditions during its lifecycle and so knowing where they all went and what they all drilled is a near impossibility.

There are 2 main ways that I prefer to browse expeditions and find relevant data. The first is to download the latest IODP KML file, load it into Google Earth and fly around the world’s oceans. Each data point has associated expedition data and it’s a great way to find the extent of IODP exploration in the area.  

IODP drillsites plotted in Google Earth from KML download
The second way is to use the Scientific Earth Drilling Information Service (SEDIS) database. The SEDIS database is a web-based search tool designed to increase the accessibility of IODP data through various search functions. These include searching for datasets by keyword/lithology, geographic area, map coordinates and date. It is also possible to search IODP publications for keywords to find references to related lithologies, structures or fields of research etc.

Searching for lithologies in the SEDIS database

2. MGDS

IODP is not the only source of publicly available petrophysics data. The Marine Geoscience Data System (MGDS) provides a service for free public access to marine geoscience research data. Since 2010, MGDS has also been part of the Interdisciplinary Earth Data Alliance (IEDA). This is a National Science Foundation (NSF)-funded service for solid earth geoscience data. One way of easily and visually browsing for data is by going to the IEDA data browser and searching geographical locations by changing layer information. Layer data includes:
  1. Integrated geochemistry data from PetDB, SedDB, MetPetDB, GEOROC, NavDat, USGS, and GANSEKI.
  2. MGDS Cruise Tracks - Marine geoscience research data acquired throughout the global oceans and adjoining continental margins. Very useful if you are searching for cruise data in specific locations.
  3. Geochronology - Community contributed database of U-Pb, (U-Th)/He, and Ar-Ar geochronology and thermochronology data.
  4. GMRT - Global Multi-Resolution Topography (GMRT) synthesis
  5. SESAR catalogs and preserves sample metadata profiles and operates the registry that distributes the International Geo Sample Number (IGSN).
  6. Seismic Data - Marine seismic data from active source studies conducted for academic research are managed through the Academic Seismic Portals (ASP) of LDEO and UTIG.
Seismic data lines shot in the Caribbean region and stored in the IEDA database.
Another way of visualising the vast quantities of data stored in the MGDS is through the GeoMapApp. GeoMapApp is free software that, once downloaded, provides a map-based application for browsing, visualizing and analysing data. With a huge number of options and functionality, you may want to make use of the Youtube tutorials to get the most out of the software.

Ice flow velocity, plate velocity vector data, and seafloor crustal age data available inside GeoMapApp

3. NDR

For petrophysical data that centres on the UK, the UK Oil and Gas authority released this year 130 terabytes of data in the form of the new National Data Repository (NDR). This data comes in the form of more than 12,500 wells, 5000 seismic surveys, and 3000 pipelines around the UK. Much of this data are historical, however there are also data on wells as recent as February 2019.

Even if this data is not directly applicable to your research area, finding real data on similar geological settings can be extremely useful for training both people and machines in the pursuit of learning.

Oil and gas fields with available free data inside the National Data Repository

So there it is, three ways of sourcing free and publically available petrophysics data to help answer all your marine geoscience questions. Happy searching.

Monday 15 April 2019

Applying petrophysics to geoscience challenges


Before we get into this topic, it is important to note how I am defining petrophysics. For a full explanation see my previous blog, but in a nutshell I use petrophysics as “a term to express and explain the physical responses of particular rocks and sediment types”. 

Straight off the bat it should be obvious that petrophysics and the associated insights it brings are applicable to a wide variety of geoscience challenges, particularly in the understanding of the subsurface. I would like to offer a few examples below but I should also mention that this article is not exhaustive. These are just a few example applications of data, primarily from the International Ocean Discovery Program (IODP) and its precursor programs, where downhole (well) log data is routinely collected and has been used to answer the scientific challenges of its science plan. 

Evidence for the wide applicability of petrophysical data is its role in the IODP science plan 2013-2023. The science plan is intended to guide multidisciplinary international collaboration by outlining the breadth of questions that IODP aim to tackle. Petrophysical analysis plays a part in answering all these questions covering the fields of: 
  1. Global climate past and present
  2. Deep life and the biosphere
  3. Planetary dynamics and tectonics
  4. Geohazards
Many of the strengths and applications of petrophysical data derive from the nature of its acquisition. Well log data provide an ‘investigation area’ that is relatively unique in the fields of geoscience investigation techniques. Sediment, rock and core samples and their analysis are commonplace, and investigate the nano-to-centimetre scale. Whilst seismic profiling provides basin-scale architecture, at its very best the data have a resolution of 10 metres. Log data covers the 0.1-100 metre scale area of investigation and is one of the few technologies specifically designed to do so. Added to the fact that logs capture continuous data and the in situ properties and, for these reasons, I think logging data should be more commonly utilised and considered when generating, and ground-truthing, geological models. 


Geoscientists can benefit greatly from integrating petrophysical data with their other data generated from direct sampling or observation of sediments or rocks. Petrophysics is geared towards answering why certain rock types exhibit the physical responses that they do, and in doing so producing quantifiable information on chemistry, mineralogy and fluid content (where possible). Invaluable information for sedimentology and petrology. 

The continuous downhole (or more accurately, up-hole) recording of data also makes it ideal for capturing and analysing stratigraphy, cyclicity and other trends. This again can be valuable information for all geoscience disciplines but really add value to geoscientists researching palaeoclimates, paleoenvironments and paleoceanography, including those interested in sediment source. 

Logging tools have different purposes with some aiming to characterise different aspects of the formation than others. Many tools such as electrical resistivity have deep areas of investigation (typically 1.5-2 metres). These tools can provide information on formation structure and fluid content rather than mineralogy and chemistry. Other tools can have extremely shallow depths of investigation such as the various imaging tools (millimetres). Borehole images can be used to analyse millimetre-scale textures and sedimentary structures within rocks and sediments as well as to examine fault orientation, dip and dip direction. 

Imaging tools are deployed as standard during IODP expeditions and are becoming increasingly common elsewhere. The examples of borehole images below cover most of the spectrum. The left image was recorded during IODP Expedition 364 using a slimline acoustic imager [1]. The image has clearly recorded the coarse grained nature of the granite, but has also captured two generations of intrusion and measurable fractures comparable with the corresponding core from this depth. In the right image, wireline electrical resistivity images and Logging While Drilling (LWD) images have captured cross bedding in deltaic sandstones 1600 feet (~490 metres) below the surface in an oil well in Oklahoma [2]


The discipline of petrophysics is closely related to geophysics and integration of petrophysical measurements with other types of geophysics data can be of great benefit. It could be argued that the most useful data are those generated by the sonic tool. The sonic log is a continuous, usually high-resolution record of compressional velocity along the well path [5]. These data can ground-truth seismic surveys in the area by establishing the time-depth relationship and, importantly, linking well log to seismic profile and ultimately core to seismic. Morgan et al. [3] demonstrate this in their scientific expedition drilling in the Chixculub impact crater peak-ring. Here the seismic P-wave velocity (km/s) obtained from sonic wireline logging data confirmed that the predominantly coarse-grained, granitic rocks of the peak ring were indeed characterised by the low densities and low seismic velocities suggested by geophysical models based on seismic refraction data. 

When the sonic log is combined with a density log it also becomes possible to calculate acoustic impedance (a property of rock layers and their boundaries that govern acoustic reflection coefficients). Combining these petrophysical log data allows for creation of a synthetic seismogram. Further insights can be made into seismic profiles if shear slowness logs are generated as these can advise on formation fluids. 

Access to fresh water is already one of society’s greatest challenges and will be an increasing concern in to our future. Lofi et al. [4] used a range of data from IODP Expedition 313 including lithology, 2-D seismic profiles, pore-water salinity measurements, porosity measurements, density measured from core, thorium content (from downhole spectral gamma-ray logs) and sonic velocities from downhole logs to determine the geological heterogeneities affecting groundwater exchanges on the New Jersey shelf. Their work revealed evidence for a multi-layered reservoir/aquifer where waters with very low salinities (<3 g/L) were encountered at depths below sea floor exceeding 400 m and fresh and/or brackish-water intervals alternate vertically with salty water intervals on this passive margin.


It is also worth mentioning borehole gravity surveys, and here I must admit that I am a bit out of my area of expertise. Suffice to say though they are to gravity surveys what sonic logs are to seismic surveys [5]. For more information on borehole gravity surveys and their relationship to surface gravity surveys see Martin Kennedy’s book Practical Petrophysics – Chapter 14: Geophysical Applications.

Further applications:


Downhole tools are becoming increasingly versatile with tools for magnetic susceptibility, fluid sampling, magnetisation and borehole imaging. Core-based petrophysics is also a rapidly expanding field with increasing commonality of chemical analysis such as XRF, hyperspectral imaging and near-infrared spectroscopy. Understanding of the data produced by these new tools is of increasing importance to academia and industry. 

Petrophysics and its techniques can also aid in the fields of: 
  1. Contamination
  2. Remote sensing
  3. Soil and sediment science
  4. Geochemistry
  5. Hydrology and hydrogeology
  6. Geotechnical measurements
And finally, what of geological models? Martin Kennedy makes a great point about this in his book Practical Petrophysics. I’ll let his words speak for themselves: 
“The increasing use of software to build detailed 3D geological models [of reservoirs] has meant that petrophysics has to be properly integrated with the other sub-surface disciplines. The model builder needs to know what assumptions have gone into the creation of the petrophysical property curves and the petrophysicist needs to know that their results are being used appropriately. Consequently a working knowledge of practical petrophysics is no longer just a ‘nice to have’.”
For ‘of reservoirs’ read any sedimentary basin, aquifer, impact crater, passive margin, mid-ocean ridge, obducted ophiolite, subduction zone, slow-slip zone – the opportunities are limitless. So if this has given you pause for thought and you are interested in knowing more, why not have a look into how petrophysics can benefit your science?



Friday 12 April 2019

Petrophysics is about much more than oil


“Petrophysics” is a term not widely used in academic circles (at least in my experience), but it is one that is quite extensively used within the language of the oil and gas industry. So what is petrophysics exactly and what does it mean in an academic context? The summary that I most commonly come across goes like this:


This is, in my view, a good definition. But I wanted to take it a step further by exploring a little history.

Petrophysics is a term generally linked to downhole (well) log measurements and their analysis (by petrophysicists) to evaluate rock properties. The Schlumberger brothers ran the first well log (or something close to it) in 1927 when they lowered an electric sonde down a well in Pechelbronn, Alsace, France to measure electrical resistivity. This was the first “down hole” measurement of rock properties using technology that Conrad Schlumberger had been developing since 1911. For a great history of the first well log and the road travelled by the Schlumberger brothers to start the international company we know today, see the Schlumberger website (definitely worth it for the pictures alone).

The first well log, September 5th 1927, Pechelbronn
So began the relationship between the rise of petrophysical analysis techniques and the growth of the oil and gas industry. Technological developments and new techniques have since stemmed from the needs of the industry and petrophysics remains a tool most commonly used for describing and analysing all aspects of the hydrocarbon system. In turn, this created a bias in the available technologies, with the majority of tools (at least originally) being designed for describing porous media and the quantities and nature of the fluids they contain.

The Schlumberger brothers
However. I would argue that, despite this ‘tool development’ the Schlumberger brothers are not the fathers of petrophysics. I would argue that this title belongs to Gus Archie, the author of two of the top 10 landmark papers in petrophysics and formation evaluation - including the famous Archie equation for determining water saturation [1]. The Schlumberger brothers were the first to develop and implement the technology, but it was Archie who was the first to understand the data. In his book “Practical Petrophysics”, Martin Kennedy discusses the history of the technique, stating that before Archie, petrophysical data were primarily used for qualitative interpretation of the sub-surface, such as identifying sands and sometimes distinguishing water and oil in pore space [2]. It was Archie who, in 1938, was charged by Shell's Texas-Gulf area production manager, D. B. Collins, with the task of understanding electrical logs [3]. And it was through this venture that Archie’s now-famous equation appeared in 1942 followed by the Archie’s first published use of the term “petrophysics” shortly thereafter in 1950 [4].

Gustave Erdman Archie (source)
In his 1950 paper Introduction to Petrophysics of Reservoir Rocks, Archie describes petrophysics as: “A term to express the physics of rocks. The term should be related to petrology as much as geophysics is related to geology. ‘Petrophysics’ is suggested as the term pertaining to the physics of particular rock types” [4].

It’s worth noting that this term may have been already used informally at the time, but as the first published example, I believe Archie should be credited with the definition.

So how does this definition differ? Martin Kennedy expands the definition with an explanation: “As a pure science its [petrophysics’] objective would probably be to explain why rocks have the properties they do. In particular how the relative amounts and arrangements of the minerals that comprise them determine their physical properties.” [2]

It is within this ‘why‘ that I think academic petrophysics can thrive. Petrophysics has its roots in understanding why rocks exhibit the physics that they do, and this is not limited to sands and mudstones (shales). While the majority of downhole tools are still biased toward characterising reservoir (sandstones) and cap rocks (mudstones) for hydrocarbon prospecting there are so many other useful tools, data and applications out there where petrophysical analysis can make a major contribution (more on that in the next blog).

All of the statements above are my own opinion.

Laurence Phillpot

  1. Archie, G. E., 1942. The electricalresistivity log as an aid in determining some reservoir characteristics, Trans.AIME, 146, 54–67.
  2. Kennedy, M.,2015. Practical petrophysics (Vol. 62). Elsevier.
  3. Thomas, E.C.,1992. 50th Anniversary of the Archie Equation: Archie Left More Than Just anEquation. Log Analyst May–June, 199-205.
  4. Archie, G.E.,1950. Introduction to petrophysics of reservoir rocks. AAPG bulletin, 34(5),pp.943-961.