Welcome!

Services:

I offer these services, worldwide (hybrid mode):

Consulting: geophysics and geology
Application support: Petrel Geophysics
Software development: specializing in scientific code
Management: seasoned Software Project Manager, Mentor and Coach
Innovation: reinventing seismic reservoir imaging, interpretation, characterization and monitoring

More about these service offerings later … but, first, why this name for my company??

Firda:

Firda is the Old Norse (Viking) name of the district in Norway where I was born and raised.

Firda is also the general Old Norse equivalent of the modern Norse word ‘fjord’. It is cognate with both the Scots ‘firth’, and the English ‘ford’. It can refer to any sort of coastal bay in Scandinavian languages, not just a long narrow one. The original may have been an extension of ‘folk’ as a plural to refer to a region in the same manner as ‘Nordwegen’ (literally ‘the way North’) became a region. The root names of the English counties of Norfolk and Suffolk (within the kingdom of East Anglia) evolved in the same way.

You can read more about the Old Norse history of my home district here.

Image: Firda seen from space. The shortest sailing route from Scandinavia to “The West” (Shetland, England, Iceland, America) is clearly from the Firda district.

Fun-fact: The first permanent settler on Iceland (which btw established his settlement in a place now known as “Reykjavik”) was a Viking from Firda. His name was Ingolfur Arnarson.

Image: The sea west of Firda (the North Sea) is also a prolific oil/gas region. The block you see in the center is the Troll reservoir. It is the largest oil/gas reservoir on the NCS (Norwegian Continental Shelf), which provides 11% of EU’s annual gas consumption (in 2024). You can read more about the Troll field here.

Fun-fact #1: We used the Troll seismic dataset (NH0301) for the development and benchmarking of the world’s first commercial AI/ML interactive Horizon Autotracker, now commercially available in Petrel. This was a suitable and challenging dataset, because the top reservoir horizon is both heavily faulted and also an unconformity (which means that the waveform is wildly varying laterally, and even has alternating polarity).

Fun-fact #2: I supervised an MSc student at NTNU in 2018. Her MSc project was to evaluate an early version of the ML/AI ‘IntelliTracker’ (as it was known then) using the Troll seismic dataset. You can find her MSc thesis here.
–

Image: This is the top horizon (interface) for the massive Troll reservoir zone outside Firda. The horizon is color-coded by the “Consistent Dip” seismic attribute, which I invented and patented in back in 2009. This attribute is excellent for highlighting geometric structure, and in particular the discontinuities (faults/fractures) of such structures.

Geo (Science):

My first experience with seismic data and seismic interpretation was back in the summer of 1990 when I got an internship with an energy company named “Enterprise Oil“. My job was to digitize interpreted horizons (on paper seismic sections!) which had been manually interpreted by the geologists using color pencils. Those digitized 2D intepretations would then be “gridded” (i.e. interpolated) into 3D horizon maps using a tool named “Sattlegger“. The same tool would then be used for Time-Depth conversion (using well markers, seismic velocities, etc) and “Property Modelling” (using well logs). The final steps of their exploration workflow were “Volumetric Estimates” and “Uncertainty Analysis”. The uncertainty analysis was done using “Monte Carlo” modelling/simulation. Quite simplistic these days, but it was rather sophisticated at that time, with the compute power available then (we used a VAX mainframe).

–

My MSc Thesis (June 1993) at the University of Bergen (UiB) also revolved around seismic data, wave propagation, and more generally the numerical solution of any PDE’s, using a coordinate-free notation. This was done by implementing a general TensorField API (in C++) which can handle any dimensionality, and which explicitly includes the metric tensor in all computations. The ambition was also to design the API such that it captures the full power of the general mathematical tensor notation and simultaneously enables multiple parallel back-end implementations, depending on the available compute hardware. I did this more than 20 years before Google released their now-famous TensorFlow library, which is the current workhorse for theirAI/ML workloads. My supervisors were Prof. Hans Munthe-Kaas and Prof. Magne Haveraaen. My MSc work (made possible by the great thinking of my supervisors) was very successful, and evolved into a full UiB research program, lasting for more than 15 years, and with a commercial company spin-off. The name of the research program was “Saga”, and you can read more about it here.

Images: This PDE solution to the wave equation has many interesting properties. It includes the full 4’th order Hooks tensor (containing 81 variables in 3D space, but reduces to 21 independent variables due to various symmetries). This implies that it can handle any type of anisotopy you can think of (without adding unnecessary compute overhead). It also incorporates the metric tensor, which is necessary if you would like to model wave propagation in curved spaces (this is needed e.g. if you would like to model ultra-long-wavelength earth quake waves propagating around the earth. In that case you would probably want to include a term for the Gravitational force as well). The code can also run on a variety of parallel compute back-ends just by swapping the implentation of the underlying field class. Further improvements to this code would be to include terms for Anelastic Attenuation (frequency-dependent energy absorption) and Dispersion (frequency-dependent velocities).

You can read my MSc Thesis here (sorry, Norwegian only): https://firdageo.com/uib_msc/
And, the appendix with source code (in English) is found here: https://firdageo.com/uib_msc_appendix/

(BTW: the grade of my MSc? I got Grade 1.3 at my oral defense and 1.5 for my written thesis, where 1.0 is best and 6.0 is worst. Any grade below 4.0 is a Fail)

–

After university I went straight to Schlumberger in January 1995 (just intervened by mandatory millitary service in 1994). I was first hired as “Research Scientist” at “Schlumberger Stavanger Research” (SSR) by Lars Sønneland (which invented the 3D Seismic Interpretation workstation). And the rest is history … I stayed with SLB for almost 30 years.

My Publishing:

I have a large body of scientific work, including 20+ peer-reviewed conference proceedings, journal publications, book chapters, lectures and also 20+ patent applications, and my work has more than 450 scitations. You can find a comprehensive list of my published work here: Google Scholar

–

You can find the list of all my 20+ patents here: Google Patents

–

I have also given a substantial number of public lectures, in particular many organized by EAGE (the European Association of Geophysicists and Engineers). You can find recordings of many of my lectures here: YouTube

Contact Details:

Victor Aarre Madsen
(Owner/Chief Scientist)

Email: [email protected]
LinkedIn: Victor Aarre Madsen
Phone: (+47) 481 54 077
–