Of the 4.6-billion-year geological history of Earth, lifeforms are thought to occupy a time slot only during the last 30% of this period, with the most development just in the last 15%, or some 700 million years, known as the great Phanerozoic epoch, comprising the older Paleozoic, the Mesozoic, and the more recent Cenozoic eras. It was some time during the Paleozoic era of a half billion years past that oil deposits are considered to have been formed.
These deposits of oil are invariably found in sedimentary rock formations, and for this reason they are co-associated in their combined and respective origins. The most cogent arguments stem from the thoughts of 19th century antiquarians, involving the simultaneous settling of mud and silt along with the conversion of animal and plant matter into petroleum products under anærobic conditions over many, many moons. However, over the last century we've found so much coal, gas and oil, and suspect the presence of much more potential deposits than what we've located thus far, that there ain't been enough dinosaurs and plants to give their all for even a fraction of the amounts recoverable. This would mean that the plantlife of the carboniferous period of some 250 million years ago would have soaked up an extraordinarily enormous amount of carbon dioxide from the primordial atmosphere, at far higher concentrations than that in which animal life could survive. Unless--to satisfy our argument--the partial pressure of oxygen was also appreciably higher to compensate for the stifling effects of CO2.
At least 20 years ago (1972) a retired Texaco engineer, Leon Gaucher, published what I consider one of the best answers to date on this problem--separate and distinct from the inorganic theory of cosmologist Thomas Gold of Cornell, who himself theorized that petroleum was actually manufactured in the Earth's interior through the heat-catalyzed combination of carbon and hydrogen. In the well-known water-gas shift between hydrogen and carbon dioxide to form water and carbon monoxide, Gaucher hypothesized that these reaction products generated the raw materials for the formation of hydrocarbons. Gaucher surmised that Earth's primordial atmosphere contained a predominant measure of carbon dioxide. He also thought that the necessary hydrogen was gleaned from the solar wind and from interstellar space. In my view, it's more likely that hydrogen was already present in copious quantities in Earth's primordial air mass.
The equilibrium reaction between the principal raw materials, namely carbon monoxide and hydrogen, over the still cooling surface of a primordial Earth at 500°C--the present surface temperature of Venus--would be catalyzed in the forward direction to form equally copious amounts of methane and water. And, as Earth continued to cool, the presence of volcanically generated catalysts in the hydrogen-laden reducing atmosphere would promote the synthesis of additional methane, all sorts of saturated and unsaturated hydrocarbons, as well as oxygenated compounds, with many promoted by the presence of lithospheric metallic oxides. This, in effect, would be Nature's ancient version of the Fischer-Tropsch synthesis, which human ingenuity developed circa 1923.
The remarkable feature about this versatile reaction, according to Gaucher, is that it would be initiated spontaneously and is highly exothermic--generating lots of heat, and is therefore self-sustaining. In the range of 400°C down to 180°C, as Earth continued cooling, all manner of long-chain hydrocarbons would be produced. Gaucher himself thought that such reactions would take place over millions of years, but, again, in my view, and depending on the equilibrium conditions, such complex reactions might just take thousands if not mere hundreds of years, and possibly less. In fact, under a near ideal condition where the atmosphere is well-mixed and fairly homogeneous, a catalyzed reaction at one stage of synthesis would take but moments to exothermally flare throughout Earth's entire air mass.
The most rapid acoustically-induced kinetic reaction of which I'm aware--the cystallization of supercooled liquid phosphorus--propagates at something like three kilometers per second. A reaction at this rate would translate into a conflagration of the entire atmospheric envelope in something under two hours. Such a runaway conflagration would effectively blow away the greater part of the atmosphere into space as a chemically-induced planetary nova. With a lessened atmospheric overburden acting as both a heat sink and insulating blanket, the planet would cool more rapidly, which is--for all the kinetics freaks reading this--strictly according to the Stefan-Boltzmann law, where the energy radiated is proportional to the fourth power of the temperature.
With this cooling would come the condensation of considerable water formed by the reaction between hydrogen and the oxides of carbon, and the subsequent development of the oceans and seas. Therefore, if the amount of water is in any way indicative of some fraction of the petroleum deposits, then there is an awfully large quantity of oil yet to be found, and we shouldn't have much of an energy crisis emanating from that source.
[Source: Frederic B. Jueneman - Raptures of the Deep]