How do you combine different data sources for drilling software?

In my experience, definitly data is most important thing for do engineer. Indentify data format, how to manage and quality. The first step before to introduce in drilling software. The most common is WITSML, LAS and DLIS. Is very common uses to Excel when import o export data or xml.

Common data sources in geothermal include geological, geochemistry, and geophysical data, such as seismic, well logs, core samples, formation tests, and other rock property information. Geothermal drilling data such as surface and downhole measurements like weight on bit, torque, pressure, temperature, and rate of penetration. Engineering data including design specifications, parameters, and models for the wellbore, casing, cementing, drilling fluids, hydraulics, and other aspects of the drilling system

Include both time and depth base data. Ideally if you can find/ use a software which allows you read/talk to hard copies or digital copies, like pdf formats

Depending on drilling software used there are different types of data for each drilling engineering calculation. Initial data is used as a basis for all engineering calculations and those calculations are based on mathematical models and formulas with several unknowns. For example well trajectory calculation is based on minimum curvature formulas. it uses basic triangular geometry to calculate one segment of trajectory which is actually a curve. to do this calculation the input values would be inclination and azimuth of three measured points on a plan and distance between these points. But for a drilling program to calculate the trajectory normally you would input dogleg which is in degrees per 30 ft and the distance or the measured depth

In my position of rig based ERD Advisor, the data sources were: 1. depth based: for making progress, whether making hole or running pipe and tools downhole. Data sources were the rig electronic data capture system (electronic geolograph) combined with WITSML or csv files periodically updated from downhole Logging While Drilling tools. 2. Time based: adhoc from the rig electronic data capture system for cementing operations volumes and pressures. Both depth based and time based data sets could be uploaded per shift into proprietary engineering software, which would require initial calibrating parameter inputs about the rig blocks for mechanical calculations and pumps & lines for hydraulics calculations. Calibration was key to results.

The quality and validity of the data sources are critical for "Garbage IN Garbage OUT". This involves checking for errors, inconsistencies, gaps, outliers, and biases that may affect the reliability and usability of the data. To do this, you can use a variety of methods and tools such as data cleaning, data verification, and data analysis.

Need to find out and understand the data acquisition rate ( hertz or data points per second) so you can make an apple to apple comparison. Data acquired at different data rate will show different values . Also understand and/or find out if there is any kind of data manipulation (average, cleansing, replacement, etc) from the data origin (sensor) to the data aggregation and transmission

To check consistency of data input the best way is to visualize all drilling calculations in a drilling software. A quick view by a drilling expert will help to check the data consistency.

So far data quality still depends on the specialists who run software either G&G or Drilling and check if results are valid based on their experience and skills. It's worth to say that it is not only software skills, but subject matter either which is even more important then the first ones. In other words, engineer who works with any software must understand what each value entered means and what kind of result should it be.

In the rig based ERD Advisor role, data validation was accomplished by observing the real-time trend monitoring on the LWD screen or rig drilling data system. This was supplemented by key real time measurements at connections of hook loads and rotary torque for the friction factor calculations required for trend interpretation plan v actual to avoid problems like stuck pipe or well-bore stability problems. Once a shift, or after critical operations ended, a subset of depth or time based data for detailed analysis was selected by plotting trends of the loaded input data in the proprietary software. The proprietary software enabled the data to be processed for just-in-time interpretation and validation in the quality of results obtained.

Data transformation, merging, and mapping are necessary for successful visualization of the data. Data transformation involves converting or standardizing data formats, units, scales, or coordinates for compatibility and interoperability. Python has been my biggest platform in data transformation. Excel, SQL, or Python have been used to do data merging combining data from different sources to create a unified dataset.

Make sure you keep/save the raw original data before you transform it in case you need to go back and reprocess it at a later date

There are a few options that can be used for data integration and visualization. One of them is Well View widely used in the industry as it combines, structures and visualises drilling data.

Managing and securing data sources is of the drilling project lifecycle. This activity involves organizing, storing, sharing, and protecting the data to ensure its availability, accessibility, and integrity.

Is important to recognize the latest advancements with cloud applications and the openness of the system’s new generation. Data exchange via email is very inconvenient and can easily derive in data breaches or design problems. For this, several very interesting data platforms are now available that allows a seamless combination of domain, data storing, accessing and sharing with the highest levels of security controls like SOC2.

In the ERD Advisor role, a key functional service was drilling optimization in well planning. Data sources provided by the client from offset wells in various external sources was captured by the proprietary software for the project. Calibrations for hookloads, pipe and pumps, etc., were applied before calculating & plotting depth trends of mechanical and hydraulics engineering results. Proprietary algorithms in the software enabled results, valid for purposes of 1. decision making (drillpipe or casing, etc.), or 2. problem solving (for instance Equivalent Circulating Densities v formation strength), or 3. performance improvement (from a detailed feasibility study with actionable recommendations or a specific drilling plan update).

Applying algorithms, models, or workflows that use the data to support decision-making, problem-solving, or performance improvement is also a critical step. For example, Excel, Python, or Matlab generate outputs such as parameters, indicators, or results. In geothermal, python is becoming a core skill especially as we endeavor to use Artificial Intelligence to optimize our systems.

Data collection methods such as surveys, interviews, or reviews can be used with tools like Google Forms, SurveyMonkey, or Qualtrics. Excel, PowerBI, or Python can be used to analyze and interpret the feedback.

Integration to real time decision making is where all these tools have to advance - simple metrics that summarize the intelligence over data collected is the target to achieve. Not there yet, as of now.

It's very important to combine different data sources that provide relevant and accurate information about the subsurface, the drilling equipment, and the operational parameters. This ensures the output will be quality.

Capturing, Storing, Analyzing and Validating Field Data is a solid foundation to providing potential aspects of the operations that potential improvements or efficiencies can be made. I strongly encourage all engineers to also communicate with field level professionals that may not have an engineering background. The good ones have more experience and knowledge pertaining to operations than given credit for and said experience is all based upon engineering practices that we are all still looking at today. Data can mislead your analysis if the data set isn’t complete (you don’t know what you don’t know)

In the ERD Advisor role on the rig, combining depth trend plots of interpreted drilling parameters with measured LWD correlating data enabled drilling v logging depth v formation marker visualization. The user may provide some judicious captions covering key features of the drilling optimization or incident investigation interpretation to aid swift decision making. This is especially important in multidisciplinary studies.