The starting point of the work flow is to select wells to include in the well set by using the filters in the accordion, configuring the tree map for marking or marking wells on the map.

There are links below the WF icons for Reset All Filters | Filter to Marked | Reset All Marking.

The heat maps and contour plots of the four maps can be used to visually explore the filtered and marked wells by color, heat and size from any attribute in the well header.

Using the production wells list to mark and review the selected wells of interest through log-log and cumulative plots, area and bar charts.

Display controls are available to set the measure units, time dimension (normalised time-zero or production date) and color by.

Marking wells in the producing wells, the marked wells can be built into a well list for the workflow.

Marked wells in the well set are passed to the MCMC autocast.  Autocast parameters control the minimum number of production months, outlier detection, start-stop detection and stochastic machine learning sample sizing  and number of months to forecast.

Parameters are defined by stream (Oil, Gas) and the MCMC DCA Autocast processes around 23 wells per second both streams.

Manual decline curves can be configured for individual wells and the start-stop parameters can be tuned using the ratio chart.

Marked wells in the well set are passed to the MCMC EUR data function.  Distributions of the MCMC samples can be seen for IP, Effective Decline and B factor for marked individual wells or multiple wells or the whole well set and production curves.

The scatter plot shows the cloud of generated Arps parameters which explain the production data. 

Controls are available to set the stream (Oil or Gas) or the parameters to be displayed plus the option to show or hide the machine learning parameter burn-in.

Links to further pages of MCMC sample, EUR diagnostics and Forecast Curves.

All numeric attributes from the well set are available to the function and you can normalise to any attribute (the default is EUR for Oil).  The first stage is a multi-variate linear correlation of all numeric attributes to the target attribute.

The tornado plot (centre) shows the positive and negative correlation to the target attribute.  Marking one or more attributes in the tornado plot of 'completion' related attributes are used as input to the second stage which is a principal component analysis (PCA).

The final action is to mark a well to normalise against, either in the scatter plot or the well list for use in the Type Well Profile generation.

Mark the wells you wish to include in the type well generation on the map or in the well selection list.  You can refresh the heatmap by selecting an attribute from the drop down list and clicking the refresh button.

The cloud plots and related curves on the page are refreshed for the marked wells.

The Type Well Model Controls determine if the generated Type Well Profile (TWP) is normalised (Yes, No) the data aggregation method (Average, Median, P-Hat, Percentile, Sum or None), the aggregate time scale (Elapsed Years, Date), Error Measure (Root Mean Squared Error RMSE, Mean Absolute Deviation MAD, Mean Absolute Percentage Error MAPE).

The Type Well can also be scaled for lateral length which can cerry into the  Economics.

The generated Type Wells are shown on the Cloud Plots and added to the well selection list.

Using all the forecasts in the well set including type well profiles, marks the wells to create economic cases for.  Set the econ parameters for each case run, for one or multiple forecasts for Timing (Drilling, Completion, Production), Investments (Drilling, Completion, P&A) tangible and intangible, Expenses/Taxes, (Oil, Gas, NGL, Water), Gas Production Metrics (BTU/MCFF, Shrink Factor, NGL/Gas (BBL/MCF), Water/Oil (BBL/BBL), Water/Gas (BBL/MMCF), Primary Phase for Water), Commercial Terms (Before and After PO), Gas Contract Type, and Price Adjustments to the loaded Price Deck.

Waterfall charts of 40 year AFIT Cash Flows by well or all wells and a scatter plot of IRR vs. NPV_0 provide visual clues to the best and worst performers economically in the well set by case.

A further page of Econ Diagnostics is available to drill into the details of every case and well.