Stuff

~ RIG TO COMPUTER ~

1976 | 1978 | 1981 | 1985 | 1986 | 1987 | 1988 | 1989 | 1992 | 1994 | 1998 | 1999 | 2000


1976 - Slumber-J: As a roughneck in my university summers on drilling rigs in Northern Alberta, Canada, I often saw blue trucks from Schlumberger then red trucks from Halliburton - the ones to lower logging tools into the borehole and measure underlying rock formations, the others to perform cementing or fracturing jobs and prepare a well for production. Roughnecks and tool-pushers on those rigs often came from nearby ranches to supplement their income, as farming was marginal even in those days and hourly pay on the rigs very good. Well folks, S-c-h-l-u-m-b-e-r-g-e-r was quite a handful to spell out when filling out work orders and contract sheets. Ranchers on the other hand were used, in branding cattle and for variety in symbols, to turn a letter on its side and call it “slumber-{letter}”. It’s not hard then to imagine that a phonetically correct Slumber-J (instead of Schlumberger) easily headed many a work order signed in Northern Alberta before the blue trucks rushed off to the next job... My bottom entry shows an updated scenario illustrating another transition from Rig to computeR.
Top Bottom

1978 - "Little old Lady from Calgary-a": As a student I took a computer course, just to please my teaching supervisor! The University of Calgary Computer Science Department has a bank of mainframes, complete with deck readers, card punchers, shelves for computer printouts, and a scheduling board to book your time (only hockey rinks were booked later at 2AM) on a time share system using JCL (job control language). I believe it was an IBM mainframe, or perhaps a DEC VAX, but its claim to fame was this: Jim Gosling broke into the system just to prove to Comp. Sci. that he was worthy of empolyment there, and the rest as they say is history (hint: he started Java and went onto Sun Microsystems, like many a Canadian export).

So the final project on my Comp. Sci. 101 class was to simulate a skin disease by having noughts and ones eat each other according to certain rules across a virtual flatscape (think Flatland and the Game of Life, you can google it). But try as I may I couldn't get the @#$%^&* program to work! So I go see my TA the Friday it was due, and he said to stew over it over the weekend and come in on Monday to tell him the result. So on Sunday night I hoof it over to the lab, as that was the only free time slot to run jobs iteratively until they worked. So I stood beside this little old lady by the card punch, and guess what? She was cussing about these letter O's and number Zero's that are indistinguishable to her poor eyesight (don't ask me what a retiree was doing punching cards, it takes all sorts I guess). Bingo! I realised I was careless in my Os and 0s (see it you can tell which is which just there LOL), corrected my code and voila! the printouts were ready in a flash and my TA was happy Monday morning. To say the a little old lady saved my belief in computers is a understatement.
Top Bottom

1981 - Opportunities and Operating Systems: August saw the launch of IBM’s PC, primitive as it was with 64 Kb RAM and no floppy disk! The operating system, PC-DOS that became MS-DOS outside IBM, was however on a roll (as was the IBM PC whose open specs fostered a hardware aftermarket that made all the difference with respect to Apple, that remained proprietary). How ironic then, that despite Apple and Unix launching the windowing desktop from Xerox, it was Microsoft Windows based on DOS that took over the desktop a decade later...

That summer I was in grad school, still boot-strapping a mainframe word processor with 8 in. diskettes; I remember learning to back up files early on, since an as-yet-unknown keystroke combination sorted my thesis alphabetically: and did it ever do it fast too… talk about “zero to dumb in under one second”!
Top Bottom

1985 - Arctic Islands: I spent a short summer in the High Arctic with the Geological Survey of Canada (GSC): do you know why the central airport in the Canadian high Arctic is at Resolute Bay on Cornwallis Island? Not only is it very near the magnetic pole (little one can do about that), but it's amidst a well-know perennial fog-bank in the summer (something could've been done about it, read on).

A coincidence in geology and geography made Siberia and northern Alberta very similar. Both had two petroleum provinces mid-continent in the foothills of orogenic belts, with shallow clastic and deep reef oil and gas deposits in significant amounts. And both were plains of rather higher elevation against mountain ranges with short summers and long days, and long dry winters. As Canada was not involved in the Cold War (indeed Kennedy suspected Trudeau sympathised with Krushtchev), there was a natural affinity and petroleum technology was freely exchanged across the N Pole. Then every two weeks a charter flights flew over the pole from Calgary to Novosibirsk. [For a weird twist of Cold War era geopolitics, see this item in another essay].

On the other side of the ideological divide, there used to be a friendly cat&mouse game going on between US and CDN 1, of lesser importance in the post-Cold War era [2009 update: but of renewed importance in the current Arctic geopolitics]: Canada claims sovereignty over the entire pie from the Yukon/Alaska border 2 up to the pole and down between Baffin Isl. and Greenland through Nares Straight. US claims the intervening international waters beyond the so-called 200 nautical mile Economic Exclusion Zone. To test that sovereignty, the US Navy used to ship a barge yearly through the Northwest Passage - from New York, south of Baffin Island, past Cornwallis Island and Mackenzie Delta, on to Prudhoe Bay. To further test this in the mid-fifties, the US decided to build an airbase to help with surveillance missions in the former Soviet Union. On the appointed summer, bureaucratic tie-ups and the extra organisation to ship an airport in spare parts delayed the departure of the convoy until September. Wouldn't you know that by end of month when they reached Cornwallis Island, the ice-bank had already closed across the straight as it does every year by October 3. So the convoy turned around, but only after off-loading said airport materials to pick up next year, and make the round-trip faster and cheaper - the Arctic is very dry and sparsely inhabited, so leaving gear up there is no big deal (I found many caches from forty years ago, with pristine cans of peanut butter, dried bananas... and spam).

So the US Navy went back next year and earlier in August, but that one was one cold summer where the ice never left the ground (same as ‘85 when I went up, which is why I was told this story). Having gone all that way and looking for something to show for their travails, they decided then&there to build the airport right where the gear was dumped on Cornwallis Island. Voilà! Instant airport with no rationale as to its location, other than traveling mishaps and ice - which, in their defence, is the rule in the Arctic anyway... It has been used ever since. Resolute Bay is a thriving community, used by CDN government field-parties that spend summers there as part of the claim to sovereignty (a whole program called Polar Continental Shelf Project went for almost 50 years though it's staggering nowadays amidst government cutbacks and new Arctic geopolitics). The only problem however is that Resolute Bay sits right in the middle of a perennial fog-bank, which any Inuit (Eskimo) could have informed anyone who cared to ask! So it's a ritual up there to build in ten days slack on each end of a short summer field season, just in case fog fails to lift at appointed flight times (and Murphy has a field-day on that one).

1: Q - do you know what the acronym C-D-N for Canada stands for?
A - Commonwealth Dominion of North-america back

2: itself under dispute with pending ANWR drilling, does it go perpendicular to shoreline (US claim as it give it more of Mackenzie delta) or straight up to the pole (CDN claim)? This happens on every shore, and is a classic GIS problem. back

3: in the field can be found nests made of stone, which were set up by Vikings to encourage eider geese to nest in - Vikings used to make October runs from Greenland to mid-Arctic and back on their sleek vessels to collect the eider down, after the geese had flown south and before the snow set in. Legend has it that they were once blown/forced off course and ended up over-wintering in Newfoundland 500 or so years before Christopher Columbus landed on Caribbean beaches (he had the better deal both in terms of climatic conditions and of what history recorded). back
Top Bottom

1986 - Geo/SQL: My start in geographic info systems (GIS) was at the Geological Survey of Canada: I compiled Arctic Islands geology in a database management system, and sent files to Ottawa headquarters for referencing against ESRI coverages (electronic map sheets) prepared at the federal mapping agency in Canada's capital. I went on to joint-venture with a geophysical company, who also needed such digital data capture services. I ended up working with Bill who later founded Geo/SQL, because we had both devised an AutoCAD front-end to Oracle and DB2, which was in high demand in the oil industry. I initially met people at Chevron, Gulf and Texaco in Calgary who saw an immediate need in posting wells and plotting pipeline routes (seismic was deemed to be a more difficult and later task). Bill was a surveyor who programmed a 10Kb geospatial kernel that handled re-projections of data in the AutoCAD window, and fetched data from Oracle or DB2 via structured query language (SQL).

Neither SQL nor GIS meant much to anybody as this was still pretty innovative. Open was not yet a buzzword and Unix was actually platform-dependant, so we decided to work on PCs running DOS: it came without graphics however (Windows was yet to become the standard it is today), so AutoCAD was chosen for its rich graphical interface; its scripting language AutoLisp was also open unlike, say, Intergraph on workstations that remained proprietary 4. My efforts faltered in both petroleum and mining, where the need for GIS wasn’t yet perceived in Western Canada (see Enigma). Geo/SQL went on to a turbulent history, change of hands in Colorado then Japan, and finally back to Bill who contracts to this day in Texas.

4: Bentley eventually broke the binary code and started MicroStation on PCs. Binary formats were common then (s.a. ESRI coverages as well as Autodesk and Intergraph CAD files) because of hardware restrictions that demanded files be small, and large datasets broken into tiles for adequate processing. Being binary also helped maintain proprietary formats, in days when files were deemed to hold competitive advantages - open standards and file exchange had to wait a while (see Prizm). back
Top Bottom

1987 - Windows: Landmark Graphics said it sold to Microsoft the brand name Windows, from its early graphical front-end to applications and databases called OpenWindows (note that Sun Microsystems has a documentation interface by the same name). The prefix "Open" survived as a brand name in Landmark's flagship petroleum database OpenWorks and OpenExplorer (see Enigma), until the advent of a company called OpenSpirit and a show called Oracle OpenWorld.
Top Bottom

1988 - Enigma Software: I joined a small startup company who took a hashing algorithm, originally written for the Texas School Board library program to catalogue books in that very large state. We stored on CD-ROMs and indexed for ultra-fast retrieval the entire well data set for the province of Alberta, Canada. It then added production and the same for the rest of western Canada on two more CDs, which were updated monthly and sold to oil companies by subscription - CDs were so uncommon then, that CD readers had to be leased to subscribers!

The data came from a local regulatory quirk: after two years operators had to make public all well and production data, which the Provincial government became maintained but did not distribute; a cottage industry developed in Calgary around that, and resulted in a comprehensive data set in the third largest petroleum province in the world. 150,000 wells were however not easy to store and retrieve, except by major oil companies with significant databasing facilities.

The CD-ROMs were a hit for small-to-medium companies, and they rapidly demanded a graphical user interface to the text searching front-end. I found a group of consultants called Enigma (some of whom were from Texaco and Chevron - see Geo/SQL), would have gladly added a GUI had the owner agreed to it - again, mapping and GIS were not yet in vogue! I went on to another data vendor, who also created later a mapping front-end to its data-delivery system (see Munro).
Top Bottom

1989 - Prizm Consortium: Proprietary data formats were causing oil companies and vendors to constantly read-write and re-write data from one format to another. One operator in Calgary, Canada got a data vendor and a software company to agree to a set of standards - respectively Gulf Canada, Digitech and Finder (now ChevronTexaco, IHS Energy and Schlumberger Info Services by way of Chevron, QC Data and GeoQuest respectively...). Data standards had been tried before, but they tended to reflect the old IT school - monolithic data models by programmers, for programmers that took significant efforts to maintain and update. The Prizm Consortium as it was called became the Public Petroleum Data Model Association, with a business-driven set of standards written as Oracle scripts to reflect both client and vendor needs. It exists to this day, unlike others gone with the vagaries of their respective sponsors.
Top Bottom

1992 - Metronet, DiscoveryPlace.com, etc.: Calgary was decidedly a hotbed of petroleum computing: over 500 energy companies crammed into a 10 square block of downtown with an comprehensive set of local data available from one government agency. That did not escape the attention of another agency, the then Alberta Government Telephone: they laid down optical fiber in most buildings for free, as a loss-leader to collect transaction fees among linked energy companies and data and software vendors. The bargain paid off as there rapidly emerged the largest WAN (wide-area network), coinciding with the advent of client-server hardware environment and open standard databases like Oracle.

The Metronet as that WAN was called, fostered other ventures but luck was running out. Fujitsu installed a large parallel supercomputer to help energy companies outsource very large data transaction processes, but that never flew. A precursor to portals was Discovery Place launched by another agency, the Alberta Research Council. It failed ironically because the optically wired city was actually slower in espousing the internet and its portals - what compelling reason had anyone to access abroad what already existed at home?
Top Bottom

1994 - Munro Engineering: I went on to do various projects in GIS and data management in petroleum and also utilities and defense (real-time tracking along trunk lines and shipping lanes, published in GISworld) before returning to my alma mater. Munro Engg. had a GIS front-end to data warehouses finally hired me - imagine my excitement in not only finding a like-minded company, but also meeting again the old Enigma crew! That company went on to be bought by Landmark, except for the non-petroleum part who tackled the US market – having been eaten alive by the likes of ArcView and MapInfo, the entire team ended up at AutoDesk. As Munro had http hooks well before anyone knew or cared what the internet was, thus was born AutoDesk’s internet application MapGuide (AutoDesk went on to purchase another Canadian liquidation, Vision which became Design for data management).

Meanwhile back'n gooddol' Tey-xas, petroleum Argus was killed by Landmark, who was commissioned by a British firm to create an Geographic Info System (GIS) front-end to their petroleum Oracle database - OpenExplorer is the product I supported until joining ESRI, creators of ArcView GIS used here. That product sells the most runtime ArcView for ESRI, and inspired most competitors in that space to adopt a similar front-end – most companies also demand ArcView for the ease of use in Unix and NT, and data interoperability through shape files, which became the format standard for GIS data through its open data specification.
Top Bottom

1998 - In the trenches: Implementing OpenExplorer at Landmark client sites around the world, I most often ended up working side-by-side with GeoQuest teams (see Y2K). I also started to work on software by ESRI, for whose product ArcView we would sell the most runtime licenses - little did I know where that would lead me eventually! In the meantime,  major accounts often split Landmark applications and GeoQuest data management, so project teams always worked together with clients, while marketing pitched the two companies against each other; this wasn’t always a smooth process, and it reminded of a purported event in 1914: at the beginning if World War I 5, Canadian and German soldiers apparently celebrated Christmas together above the trenches... and the next day resumed shooting each other! Metaphors aside, clients quickly saw the benefits of GIS front-ends to data warehouses, and demand for similar functionality increased radically after a slow start – the benefit of GIS in data management was taking hold of the petroleum industry’s imagination.

5: in the euphoria at the beginning, what was briefly called the gallant war was thought to be over next spring; the war to end all wars as it was later called, marked the transition from chivalresque charges of the light brigade (as in the Crimean War over 50 years earlier), to grinding trench and tank warfare: the horrendous human sacrifice, barely abated by large-scale deployment of the Red Cross (also started in Crimea), was also a dim glimmer of what would follow a mere 25 years later during World War II... back
Top Bottom

1999 - Diatomite: One of my more successful data management projects was at Aera Energy in Bakersfield CA. Just north of LA across the Grapevine (a cross range caused by a seaward bend of the San Andreas Fault), lie a forest of oil wells on sub-acre spacing, almost a hundred years in age. Oil seeps belie a field so rich, that one simply drills and finds heavy crude just light enough to produce by steam-flooding (the project consisted of improving data management processes with the intent of reducing steam injection costs, and we'd get a cut of the savings... one way to leap into the big league!).

Geologically this was a strange province, because the producing formation is diatomite, tightly packed and very prolific skeletons of microscopic size inside which living organisms decayed into petroleum, in an area of high heat heat flow (high temperatures on the rim of the Pacific ocean with active plates and volcanoes). The organic content and temperatures were so high, that petroleum was produced literally in-place, without rocks being buried very deeply (the only analogue is in Indonesia with similar heat and burial regime, and very rich jungle vegetations being cooked into oil). Such formations had high porosity (lots of space inside and in between the skeletons) but low permeability (little if any connection between the pores in the skeletons, tightly packed as they turned into rock). There was also no deep burial of the oil-producing formations, and there is no pressure of hydrostatic recharge (pressure of water trying to reach the surface, like in hot springs along the San Andreas Fault) - the classical configuration of oilfields saw gas pushed on top of oil, itself on top of water. Here gas if there was any simply evaporated through surface cracks, water if there was any escaped through the deep faults, and oil simply sat in the tight formations from which it could not easily escape (low permeability).

Oil was also so prolific, that there was no real need for any geology or geophysics to place producing wells. One simply drilled, completed, pumped and logged often enough to track oil production. Logging here consisted however of lowering electric tools to follow the drop of the top of the oil, not the rise of the gas/oil of oil/water contact as customary. This is because there was no formation pressure pushing the oil up (no hydrostatic recharge) - oil simply drained out of the formation, like water out of a bathtub (oil and gas more commonly come out the top like froth out of a champagne bottle). In fact no-one actually really knew what happened in the rock formation above the oil as it drained! The emptied top of the formation was called air-sand, a term found nowhere else in the annals of oil production. It was real enough though: if one pumped oil out too quickly, the formation near the surface collapsed very slightly as oil no longer propped up the rock formation, and the pump jacks got jammed in very slight depressions (in fact so shallow that they're measured with laser-ranging devices used nearby to track earth movements along the San Andreas Fault). So little was needed to be known of the subsurface to produce oil over the first 75 years, that when steam was injected in a second phase of enhanced recovery (common in oilfields exceeding 50 years of age), steam did not go where it was supposed to go... and oil recovery was only marginally improved (hence my project to improve production through better data, in exchange for a cut in the savings on steam injection).
Top Bottom

2000 - Y2K: Few periods in computing history saw so much effort put into and so little learned from, than the months running up to the year 2000. It started as a fair enough concern that old computer code, written decades ago to save space and thus entering dates as two (98) rather than four digits (1998), would not know how to interpret the year when it turned to 00, as 2000 or 1900? Snake-oil peddlers and honest managers alike got caught up in a maelstrom that went over the new millennium without even a pop, except for the noise created by the self-styled fear of Y2K.

Needless to say my employer got caught up in the act, as I was involved in data management projects where dates were not insignificant. That gave me an opportunity to move from Houston to London, and have our baby daughter get closer to grandparents in England and France! So I end up at BP in Sunbury, its old research center dating back to World War I, and now a hub of global data management from Azerbaijan through Brazil and Vietnam to Sakhalin. We had three months to rationalise their software, reduce its numbers to manageable levels as management changed its tune: it turned silos of exploration, production, finance and research (each with their jealously-guarded turf), into asset teams responsible for oil provinces from cradle to grave (which meant integrating all business processes); this is not unlike switching car production from assembly-lines at Ford for Model Ts, to teams responsible for each individual car at Volvo over half a century later.

On the shop floor where data is acquired, managed and handed to the geologists, engineers and accountants, Landmark and GeoQuest shared the turf - one managed data and the other the applications that turned data into information. Lots of toing-and-froing ensued with a year-end deadline looming, because one could never tell what might happen with those chopped-off dates! Both companies were pitched against each other in sales and marketing, as part of the two largest oilfield service firms, Halliburton born in Oklahoma and Schlumberger near Paris at about the time Model T Fords were built (my top entry illustrates an oilfield scenario repeated ever since)... Meanwhile back at the ranch (or the cottage) we simply had a job to get done in less than three months... So I helped forge a love-hate relationship with my 'competitors', who in time turned into camaraderie and cooperation for the common good of s/he who paid our bills (I had cut my teeth at Mobil in Dallas on a similar project almost five years earlier, with the deadline looming... that of the future ExxonMobil merger!). That effort bore fruit in my latest job, when both companies called on me to help them build GIS front-ends for their enterprise database offerings (rumor had it in fact that I joined ESRI, just so I could work above-board with both!).

Indicative of the fallacy of Y2K is that while the project came in on time and budget, a few items impossible to finish by year-end (as agreed to by all) quietly trickled on past the new millennium with neither (wo)man nor machine missing a beat...
Top