Quartz Energy Geosteer Assist - Gamma Ray-Assisted Tops Identification

Background:

Our client is hoping to model underground strata for to help in the automation of well drilling technologies.  In this challenge, we’ll be using information extracted from vertical wells and instrumentation data to make inferences about the local environment in a horizontal wellbore.  We have provided some documents in Google Drive to help explain what you need to do here

You should open the H&V Diagram.pptx first.  Quartz Energy drills two types of wells: vertical wells and horizontal wells.  The vertical wells go (generally) straight down.   The horizontal wells are actually L-shaped.  They go basically down to a certain depth and then they turn parallel to the surface.  What they are in fact doing is attempting to stay in a certain underground formation or strata which contains hydrocarbons.  But staying in these formations 7,000 or 8,000 feet underneath the ground is a delicate procedure and our client needs to understand the underlying geology as completely as possible.  One of the techniques that our clients use to understand the geology is use the measurements from vertical wells in the same oil fields to make inferences about the geology in nearby wellbores.    In our first and second challenges we developed a java-command line application that uses Spline Interpolation and Inverse Distance Weighting algorithms to make location-based inferences about the underlying strata.  This is a good start but now we want to include instrumentation data readings into the strata estimates.  In this challenge, we’ll be taking this effort to the next step.  Gamma ray reading provide an accurate means of detecting changes in geological formations.  The current application creates a report which shows the relative position of each of the strata relative to the wellbore at each recorded interval.  We’ll be adding some columns to the existing reports to reflect the influence of the gamma ray readings.

The following files have been used to produce the location-based strata estimates:

• TC - Vertical Well Survey Data -- these provides the coordinates for the Vertica Wells

• TC - Vertica Well Strata Data -- these provide the depth coordinates for the formations/strata

• TC - Horizontal Wells - this provides the surface coordinates and trajectory information for horizontal wells.

The previous challenge used 4 of the Vertical wells that are closest to the Horizontal Wells to make some guesses/estimates about the strata beneath the surface location of the Horizontal wells. Our client has given us formation data from about 50 vertical wells.  Using this data we're going to create interpolated values for the following strata based on the horizontal wells surface locations:

Strata1
Strata2
Strata3
Strata4
Strata5
Strata6
Strata7
Strata8
Strata9
Strata10

Now in addition to the location-based data we’re providing gamma ray responsivity data for our Vertical and Horizontal Wells as well.  The following documents are being provided as follows:

• Horizontal Well Gamma Ray Data V3.xlsx - Gamma Ray readings at 1 sec intervals for our 5 vertical wells split across two worksheets.  Bit Depth in the Gamma Ray data corresponds to Measured Depth in the TC - Vertical Wells Survey V3.xlsx file.

• vertical_wells_V3_gamma_ray_data.zip - A set of csv files that list depths and Gamma Ray reading for all the Vertical Wells in the data set.

The current application produces the following output:  

1. A plot which shows the Vertical Wells by surface location (XPATH_ADJ, YPATH_ADJ for Vertical Wells.   The plot should also follow the wellbore location track for X_ADJ, Y_ADJ for Horizontal Wells.  In other words we should see a line which shows the track for each  of the horizontal wells in addition to the single pixels  that will represent the Vertical Wells.  

2. A formation report (xlsx or csv format) which lists the each of the Horizontal Wellbore Ids and the depth that each formation listed above based on each given measurement  of the Horizontal Well, wells were used for interpolation, the distances involved and how the depth measurements were calculated for each step based on the current location of the wellbore.   This report will be pretty large.  It will contain 1750 rows just for Well Id 8225 ( since we have 175 rows of measurements for this well and  10 formations that we’re currently tracking).  And of course, there are 4 other Horizontal Wells to consider as well.  The current “detail” report also has a row for each relevant  Vertical Well that is being used for interpolation.    This is type of detail is preferred.

3. A relative position report (xlsx or csv format) which displays the distance of the wellbore from each of the strata or the position within a certain strata at the current location of the wellbore.  This should be repeated for each measured location provided in TC -  Horizontal Wells - V2.xlsx.  Note that as the Horizontal location of the Wellbore changes the strata interpolation estimates might have to be revised based on current position.   This report will be pretty large.  It will contain 1750 rows just for Well Id 8225 ( since we have 175 rows of measurements for this well and  10 formations that we’re currently tracking).  And of course, there are 4 other Horizontal Wells to consider as well.

New Requirements

1. formation.csv:  Please add the following information to the Formation Report:

a. Add measured depth (MD) to the report pulled from the Horizontal Well data.   The engineers and geologist like to see both True Vertical Depth and Measured Depth.

b. Update the column header “Formation Depth” to “Interpolated Depth”.

c. Add Vertical GR column which list Gamma Ray readings for each of the 4 wells used for Location-based interpolation

d. Add Expected GR value column - use closest Vertica Wells' GR values to make an estimate of the GR response value should be at the depth indicated.

e. Add Horizontal GR Value column:  show the GR value for the current Bit Depth if it is relevant.  The horizontal GR data will only be relevant if Bit Depth is near the strata.

f. Add GR Adjusted Depth column:  use the horizontal GR value to make an inference about the relevant strata depth and show that value.

2.  relative position.csv:  

a. Add measured depth (MD) to the report pulled from the Horizontal Well data.  The engineers and geologist like to see both True Vertical Depth and Measured Depth.

b. Add Horizontal GR Value column:  show the GR value for the current Bit Depth if it is relevant.  The horizontal GR data will only be relevant if Bit Depth is near the strata.

c. Add Inc Column (for Inclination) and list the value from the Horizontal GM data.

d. Add GR Adjusted Depth column:  use the horizontal well's GR value to make an inference about the relevant strata depth and show that value.   The horizontal well GR data will only be relevant if Bit Depth is near the strata.

e. Add GR Adjustment Distance column: Subtract original strata estimate from adjusted value based c above.

3.  Gamma Ray Depth Inference:  We recently ran a Marathon Match which generated code which trains a model to detect changes in strata using both gamma ray and resistivity logs.  Our client’s data scientists believe that this code might be useful/portable to this problem domain, though of course we’re only providing gamma ray not resistivity info for this challenge.  The link to the code and solution documentation will be provided in the forums attached to this challenge.

4.  IMPORTANT NOTE - Gamma ray information collected when a drill bit is moving horizontally may not display the same response values as when the drill is moving vertically although changes in strata are detectable.  The inclination value provided in the GR information is significant.

Technologies

Java 8

Final Submission Guidelines