It may be conventional wisdom to assume that outstanding oil and gas discoveries can no longer be made anywhere except in "elephant hunting ground," that is, in exotic places heretofore untapped or underexplored such as Kazakhstan, inner Mongolia or the Mackenzie Delta. One may suspect impressive finds can only be made these days by applying massive new computing power and the latest downhole technologies. As the old song goes, "'tain't necessarily so." A study of large finds made during the past century indicates that recent (1986-95) billion-barrel discoveries are typically made within 20 kilometers of a known hydrocarbon accumulation. What's more, they have usually been found by such routine petroleum activities as redrilling a previous operator's dry hole, drilling deeper within known fields, or drilling in deeper water. The industry is well aware of the large potential of drilling in deeper water-but this article will focus on the unrecognized success of redrilling previously drilled structures. Incredibly, the largest oil discovery and largest gas discovery made in the world during the 1990s were both redrills of old dry holes! On a smaller reserve scale, these same patterns seem to repeat themselves in the domestic U.S. petroleum provinces. From studying a major's more than 100-well onshore exploration drilling program during a seven-year period, we found that more than 95% of the present worth added from the program came from only two wells. One was a redrill of a previously tested zone, and both were deeper-pool wildcats. Given these repeating patterns, the systematic search for missed pay and deeper-pool targets should be a large part of exploration at any scale, domestic or international. Such an approach gives smaller companies an alternative to the high drilling and acreage costs associated with deep water. Redrilling an old dry hole necessitates the correct understanding of that well. This typically involves integrating rock, log and test information into a multidisciplinary interpretation of why the old well should have been a discovery. If done correctly, this will combine the best of both worlds: large reserve potential, yet low risk. Quantifying closeology It turns out the best place to look for oil is near a field. Oilmen have intuited this for years-they jokingly call it closeology-but now we have some hard evidence. We constructed a database of more than 45,000 wildcat wells and more than 12,000 fields in the world (excluding the U.S., former Soviet Union and onshore China). This was manipulated to replay the world's drilling history as it occurred in each of the past 120 years. This made it possible to quantify closeology, or how far away the closest existing hydrocarbon accumulation was at the time of the discovery. An estimate of wildcat economic chance factor versus distance to the closest known discovery is shown in Figure 1. This graph can serve as a practical benchmark to guide in chance factor selection. A wildcat drilled 40 kilometers from the closest known discovery historically has only a 5% chance of turning into an economic success. This data also showed that 80% of the reserves from large finds (those larger than 100 million barrels of oil equivalent) were discovered within 20 kilometers of a known discovery (Figure 2). Note also that almost all of the billion-barrel finds were also discovered within 20 kilometers of a known discovery. From this it can be concluded that drilling close to known hydrocarbons not only gives a realistic chance of success, but is also the best place to look for large reserves. Conversely, "frontier" exploration (drilling greater than 100 kilometers from a known discovery) historically has less than a 1% chance of economic success. Walking away The location and play type of the 11 recent (1986-95) billion-barrel fields is shown in Figure 3. Two of these were redrills of old dry holes that had actually drilled and unsuccessfully tested reservoirs containing billions of barrels. Cusiana Field in Colombia's Llanos Basin is likely the largest oil discovery made during the 1990s. Figure 4 shows a predrill and postdrill map of the area where this happened. Two very large structures (more than 20,000 acres) both tested small amounts of oil in the 1970s. After spending large amounts of money and encountering difficult drilling problems, Colombia's national oil company, Ecopetrol, abandoned the wells as uneconomic. But as we now know, it wasn't necessarily so, for what happened next is one of the more remarkable stories in this business. Dallas-based Triton Oil licensed this acreage in the early 1980s. On three separate occasions (1982, 1986 and 1990), senior management talked to scores of companies, trying to find partners to help satisfy drilling commitments. BP, Total and Triton finally redrilled the Cusiana structure successfully in 1992. Development wells have tested up to 35,000 barrels per day. The partners' nearby Cupiagua Field (also drilled and abandoned in the 1970s) potentially has a cumulative oil column more than 5,000 feet thick. This reminds all of us that great care should be taken in reexamining "dry" holes on large structures. The industry could not recognize these large fields even after they had been drilled multiple times. Triton's 1990 farmout effort-turned down by about 70 companies-was made after BP had tested gas and condensate from the Cusiana #1 drilled in 1988. How and why did this happen? Ecopetrol's early tests had massively damaged the reservoirs, inhibiting the tested rates of hydrocarbons. Fewer and fewer companies spend time these days studying and integrating rock data (cuttings and cores) into their well interpretations. At Cusiana and Cupiagua, this type of back-to-basics rock data was the key to understanding the true potential of the reservoir. It would have shown the industry, even back in the 1970s, that a string of giants was waiting to be found here. Deeper-pool wildcats Four of the world's most recent 11 discoveries that are more than 1 billion barrels in size were made by drilling deeper within existing fields, hardly a high-tech concept (Figure 3). This is really a variation on the missed pay concept already described, except these deeper, productive zones had not yet been penetrated by the older wells. The Arco Wiriagar Deep gas discovery in Indonesia is illustrated in Figure 5. It is instructive to identify what data was available before this 7-trillion-cubic foot discovery was made. The Dutch mapped a large surface anticline in the 1930s, on which they drilled a number of dry holes. It was not until 1981 that Conoco made a small, shallow Miocene oil discovery on this anticline. Offshore to the south, Occidental Petroleum found two uneconomic gas fields from deeper, Jurassic clastic reservoirs. This deeper, clastic section was first penetrated on the large surface anticline by the drilling of the Arco Wiriagar Deep #1 well. It tested 30 million cubic feet of gas per day. To the east of the Wiriagar Deep discovery, Total had drilled a dry hole on another large surface anticline. Although it had strong Permian gas shows when drilled, a test produced only some water. This feature was redrilled in 1996, resulting in the Vorwata Field (the world's largest gas discovery made in the 1990s). The test of water is now thought to be simply drilling fluid, making this another "missed pay" example. In another example, a major oil company's discoveries were studied during a 10-year period. In that time, many hundreds of millions of drilling dollars were spent, resulting in seven finds larger than 100 million barrels of oil equivalent (BOE) recoverable. Total gross reserves of these discoveries will top 4 billion BOE. Six of these discoveries were redrills of a previous operator's "dry" holes. Many of these actually had multiple dry holes on them before their discovery. Only one of the seven large discoveries did not have a well through the reservoir yet. It was, however, a deeper-pool wildcat, with numerous shallow wells and production. All of the discoveries were made within 20 kilometers of a known hydrocarbon accumulation. U.S. applications Are recent large discoveries in the U.S. often deeper-pool or redrill successes? Can we see the patterns described here repeated closer to home, yet on a smaller reserve scale? We think so. A study of a major oil company's 100-well exploration program (onshore, non-Gulf of Mexico) during a seven-year period shows some striking similarities. Although wildcat success rates were about 15%, almost all the present value of the program (several hundred million dollars) came from only two wildcat wells. Both of these wells were deeper-pool wildcats drilled on old, shallow, producing acreage. One of the two wells was also a redrill of a "dry" hole. The larger of the two discoveries occurred below the Wilburton Field in eastern Oklahoma, where shallow, thrusted Pennsylvanian Spiro production was discovered in 1960. In the entire Arkoma Basin the oldest and deepest production was from the Silurian Hunton level. At least one unsuccessful well had been drilled to that level in Wilburton Field, apparently condemning any deep potential. In the mid-1980s, however, this was proved not necessarily so. A geophysicist who had worked the Permian Basin of West Texas was reassigned to this area of Oklahoma. He mapped a large, untested deep fault block below a portion of Wilburton, and pointed out its similarity to the prolific Ellenburger fields found throughout western Texas. A "tail" was added onto a development well to test this deeper block. This very inexpensive deeper-pool test resulted in the discovery of the Wilburton Arbuckle (Ellenburger) Field-with estimated recoverable reserves of 350 billion cubic feet of gas! Within two years, field production rose from about 1.5 Bcf per month to more than 10 Bcf. This major oil company's other important discovery was made below the giant, shallow (3,500 feet) Hugoton Field in Kansas. In the early 1980s, an anomalous structure was mapped on this company's acreage using shallow well control. A deeper-pool wildcat was drilled to test Mississippian potential below this shallow structure. The well did find some porosity, but tested mostly water with minor amounts of oil. In the late 1980s, another wildcat well was drilled updip of the oil show, and this resulted in the discovery of Big Bow Field-about 4.5 million barrels recoverable. Well payouts were measured in weeks, not months or years, with drilling depths to the Mississippian only about 5,500 feet. Wells flowed 300 barrels per day. After some additional drilling, the company realized that this first "dry" hole was likely within the field boundary, so it was reentered. It has since become one of the best wells in the field. Operators make only one billion-barrel find per year in the world, on average. Given this rarity, it seems somewhat surprising that most of the world's recent billion-barrel finds come not from far-off places, but in familiar basins, often from new wells on old structures. This likely reflects a number of important underlying trends. First, we know where the world's petroliferous basins are. During the past 30 years, true frontier successes are increasingly rare (if nonexistent). With each passing decade, frontier exploration has become an increasingly poor business endeavor for oil companies. Second, except in deepwater plays or for political reasons, most of the large structures in these petroliferous basins have already been drilled. Some that were initially thought to be dry, did indeed contain hydrocarbons. It follows that a systematic reexamination of "dry" large structures of the world should be a major part of any exploration program (domestic or international). These types of opportunities would likely be much lower-risk than traditional exploration, given the well data already present on the structures. Third, the industry has done a surprisingly poor job of understanding what is in the wildcat wells that it drills, even when a giant accumulation is present. This typically requires a thorough understanding of the rocks that have been drilled. Few companies attempt to incorporate rock data into petrophysical and well-testing analysis any more-yet this rock data is often the key to seeing hydrocarbons or plugging the well. Finally, there is a great deal of missed pay within existing fields, both in deeper pays and within existing wellbores. Obvious producing zones gain all of the attention. Subtler, "dry" zones within existing wells could benefit greatly from a multidisciplinary reexamination of the data. Exploration staff members need to be involved in the systematic search for deeper-pool candidates. Worldwide, discovered volumes of hydrocarbons are decreasing at an alarming rate (Figure 6). As an exploration province the world is rapidly maturing. Each year, it gets harder to find the new low-cost hydrocarbons that keep companies healthy and make Wall Street happy. As discovered volumes dwindle, companies cannot afford to ignore strong historical patterns of success associated with redrilling old structures. They cannot afford to improperly interpret wells they themselves have drilled. Those that learn from these historical patterns will be amply rewarded. Mark Montie was a geophysicist with Arco for 18 years. In late 1999, he co-founded Dallas-based Mustang Resources LLC, whose primary goal is to help companies look for low-risk, high-potential, bypassed pays in dry holes and fields in the U.S. and overseas. He thanks Bob and John Sneider for showing him the importance of missed pay around the world; Jeremy Benton for his groundbreaking insights into the importance of distance in exploration success; and Arco for permission to publish the data and many of the figures used in this article. Contact Montie at markmontie@yahoo.com.