PRAKLA-SEISMOS Report 4 / 1971 | ||||
50 Jahre angewandte Seismik | ||||
Vortrag von Th. Krey auf der EAEG-Tagung 1971 In Hannover Viele in diesem Vortrag mitgeteilten Tatsachen stammen von George Elliot Sweet. Meiner eigenen Erfahrung wurden mündliche Mitteilungen älterer Kollegen hinzugefügt. | ||||
Wenn wir nach dem Geburtstag einer angewandten Wissenschaft oder Technik fragen, ergeben sich gewisse Probleme. Sollten wir uns auf den Zeitpunkt beziehen, an dem die ersten Ideen auftauchten, die dann zu einer neuen Technik führten? Sollten wir die ersten tastenden Versuche gelten lassen, auch wenn sie noch keine wirtschaftlichen Konsequenzen hatten? Oder welche anderen Gesichtspunkte sollten als maßgebend betrachtet werden? Fortschritte in der Interpretation Die Geschwindigkeit als eine Funktion der Raumkoordinaten wurde bereits - angenähert oder präzise berücksichtigt vor mehr als 20 bis 30 Jahren durch die Konstruktion von gekrümmten Strahlen und Weilenfronten. Um die Konstruktion der Reflexionshorizonte zu vereinfachen, wurde mittels zwei Arten von Korrekturen eine gewisse Vor-Bearbeitung der Daten eingeführt. Durch statische Korrekturen wurde der Einfluß der Verwitterungsschicht und der Topographie berücksichtigt, und durch dynamische Korrekturen wurden der Schußpunkt und Beobachtungspunkt in eine Position in der Mitte zwischen ihnen gebracht, so daß abgehender und reflektierter Strahl zusammenfielen. Diese zwei Korrekturen konnten bereits in den Seismogrammen vor mehr als 15 Jahren in den früher erwähnten Analog-Rechenzentren ausgeführt werden. Zu dieser Zeit wurden darüber hinaus, z. B. bei PRAKLA-SEISMOS, Digitalrechenprogramme entwickelt, um auf die korrekte Position der Reflexionshorizonte aus den Reflexionszeiten schließen zu können. "Fallgruben" bei der Interpretation von Reflexionsseismogrammen waren ebenfalls bereits vor mehr als 20 Jahren bemerkt worden, wie z. B. Multiple, Diffraktionen und reflektierte Refraktionen. Bei der Behandlung der Diffraktionen wird vornehmlich das Huygens'sche Prinzip angewandt. Es ist sehr interessant, daß dieses Prinzip bei den neuesten Erfolgen in der Datenverarbeitung -bei der Herstellung von migrierten Seismogrammprofilen -eine wesentliche Rolle spielt. Mit diesem Prozeß, der Rechner mit besonders großen Kernspeichern und Rechengeschwindigkeiten erfordert, werden die seismischen Reflexionssignale in die Position im Untergrund gebracht, von wo sie herkommen. Ein gewisser Anreiz für die Einführung der migrierten Seismogrammprofile wurde durch die Holographie gegeben, obwohl dieser Analogievergleich nicht ganz richtig ist. Es gibt noch eine andere Kennzeichnung der Migration von Seismogrammprofilen; wir können nämlich sagen, daß bei dieser Neuerung die Prozesse des "Anreißens" der Reflexionen und die Migration in der Reihenfolge umgekehrt sind. Dieses "Anreißen von Reflexionen " kann ebenfalls automatisch ausgeführt werden, muß aber noch verbessert werden. Ein weiterer Fortschritt im Vergleich zu migrierten Seismogrammprofilen ist die Migrations-Stapel-Operation. Bei diesem Prozeß wird die Migration vor oder gemeinsam mit dem Stapeln ausgeführt. Wir können sagen, daß die Explorationsseismik seit ihrer Geburt eine gewisse Perfektion erreicht hat, soweit dies den Nachweis und die Lokalisierung von Inhomogenitäten betrifft, die seismische Signale abstrahlen können. Mit Aussagen üben den physikalischchemischen Charakter dieser Inhomogenitäten stehen wir jedoch ganz am Anfang. Die Beziehungen zwischen den Reflexionsschwingungen und der Fazies der Schichten oder der Gas-, 01-oder Wasser-Füllung ihrer Poren sind noch immer ziemlich unbekannt. Hier muß noch eine Menge Entwicklungsarbeit getan werden. Ausblick Meine Damen und Herren, ich glaube, daß Ihnen dieser Bericht gezeigt hat, wie die Explorationsseismik unter dem Druck der ständig wachsenden Anforderungen der Erdölindustrie und anderer Klienten bemüht war, sich in den Bereichen anderer Wissenschaften umzusehen, um befruchtende Ideen für ihre eigene Entwicklung aufzugreifen. Wir müssen aber auch feststellen, daß Ideen in unserer Industrie geboren worden sind, die kein Analogon in anderen Techniken haben, z. B. die Mehrfachüberdeckung. Andererseits sind gewisse durch die Explorationsseismik entwickelte Ideen auch von anderen Wissenschaften oder Techniken übernommen worden, wie z. B. das "Pattern-Registrieren", das in der "normalen" Seismik erst vor kurzem unter dem Namen "Array" eingeführt worden ist. Ich glaube aber auch, daß die Explorationsseismik der Computer-Industrie wertvolle Impulse gegeben hat. Alles in allem meine ich, daß wir auf die vergangenen 50 Jahre mit Stolz zurückblicken können. Die bisherigen Errungenschaften verpflichten uns, unsere Forschungen fortzusetzen und den Stand unserer Wissenschaft immer wieder zu überprüfen, um den vielen Notwendigkeiten und Problemen, die noch vorhanden sind, nachzugehen. |
EAEG-Meeting June 1971 50 years of exploration seismies A paper by Th. Krey, given at the EAEG-Meeting 1971 in Hannover. Much of the facts presented in this paper are extracted from George Elliot Sweet. Furthermore oral communications by older colleagues have been added to my own experience. If we ask for the birthday of an applied science or technique certain problems arise. Should we stickt0 the time when the first ideas leading to the new technique appeared? Should we stick to the first tentative tests, even if they did not have any economic consequences? Or which other points of view are decisive? I feel that for exploration seismics there is a date which can most suitably be regarded as being its birthday. I think of the foundation of SEISMOS GMBH on the 4th of April 1921 which is about 50 years ago, because at that time exploration seismics started to become an economical industry. Now, ladies and gentlemen, we all know that each birth has some preceding events, and this is also valid for applied seismics. But I shall not stay on ideas and tests which had no practical consequences, though they may be very interesting, as e. g. the activity of the Geological Engineering Company with Hasemann, Karcher. Eckardt and McCollum, in 1920. In fact they succeeded in observing seismic reflections as early as June 1921 as reported by W. B. Robinson in has paper "History of Geophysical Exploration" in The Time Break, Spring 1971. Very often, the fundamental idea for a new technique is created by relating two or more branches of sciences to each other. With exploration seismics, mining surveying, an essential branch of mining sciences and geophysics were combined in the brains of Mintrop, and this combination led to the origin of the new technique which has become so important in the meantime. From mining surveying Mintrop knew the problems which required a solution and from geophysics in his studies at Wiechert's observatory, and by observations which he made during World War I, he found a solution. His ideas were patented in 1917 and in 1919 and were transferred into practical activity by Mintrop's wellknown vitality. As most of you certainly know Mintrop successfully applied the refraction seismic method in many parts of the world during the twenties. But his main activity was not in Germany or in Europe but chiefly in the United States of America and in Mexico. He prospected for saltdomes and found them and various oil fields connected to these domes resulted. But other kinds of geological structures were also succesfully examined by him. The party chief of the first seismic party successful in finding an oil bearing saltdome, the Orchard Dome, is sitting among uso It is Dr. Otto Geussenhainer, 79 years old now. Starting from about 1925 American societies too became active in refraction seismics. I specially should like to mention the Geophysical Research Corporation and the names of Everett, Oe Golyer and I. C. Karcher, furthermore Marland in Ponca City and his geophysical division including such famous names as e. g. Burton, McCollum, and Frank Rieber. Reflection Seismics and the Development of Recording Technique Today we may ask: Why did Mintrop practically neglect reflection seismics? I feel the reason is that the time was not yet technically prepared for the use of this exploration tool which is so important today. There were only some tentative tests by Mintrop and his scholars and as we know from today's experience these tests were not in a favourable area. But neither did the first interesting tests in the United States of America lead to great practical consequences. This can weil be understood from our present knowledge. Recognizing refraction events, which were nearly exclusively first arrivals according to the usage of that time, was not very difficult if the ambient noise was not too large, and if the natural frequency of the geophone was low enough. With Mintrop's geophone this frequency was approximately 10 cps. The signal-to-shot-noise-ratio was therefore no important problem for refraction seismies. With reflection seismics desired signals are mixed with a lot of other ground movements originating from the shot. Therefore it was very soon realized that it would be necessary to enhance the reflection signals by special means in order to discriminate between them and the undesired ground movements. This resulted in the invention of recording centrally a multiplicity of traces by applying the advances which electrotechnics and especially electronics had gone through. In this context I specially mention the development of valves and electronic amplifiers during the twenties. The incorporation of a new branch of science, that is electronics, into exploration seismics, thus led to a decisive advance. According to today's thinking the success which was reached by central recording was the possibility of making visual correlation from trace to trace. In this context I should like to mention again the Geophysical Research Corporation. They built the first electronic amplifiers by 1925 to my knowledge and a four trace recorder by 1928. By the way, equipment of that early time, including a Mintrop station and a Petty electronic amplifier are to be seen in the museum for applied geophysics in Houston. Though the first successful commercial reflection seismic survey had already been carried out in 1927 by Duncan, reflection seismics became a widespread technique, replacing most of the refraction parties, only after 1930, the year when GSI was founded and when the Society of Exploration Geophysicists came into being with de Golyer and Mintrop as the first Honorary Members. In Germany the development of reflection seismic equipment was somewhat younger. Here it is mainly due to Friedrich Trappe, Waldemar Zettel, who is sitting among us, and Hubert Lückerath, who was the first reflection party chief in Germany. The first commercial reflection party commenced in September 1934, the client being a coal mining company. Central recording was indeed an important advance and corresponded to the requirements of the oil and mining industries. It was the precondition for many further steps of development which are still to be mentioned. By the way, central recording was introduced in normal scientific seismics only quite recently with the introduction of arrays. During the thirties itwas realized that central recording alone was not sufficient in many areas in order to get reflections from the desired depths. But the way taken to use electronic amplifiers made possible the incorporation of all advances in electronics into exploration seismics in a reasonable manner. Here in the first place gain control has to be mentioned. Especially by automatic gain control the necessary number of shots could be essentially reduced. By increasing the number of amplifiers per recording truck and by advances in shot hole drilling it became possible to startwith continuous profiling. By electric filters the separation of reflection and noise waves could be carried out in the frequency domain. But very soon the attempt was also made to separate these different kinds of waves in the wave number domain too, i. e. by using multiple geophones and shots. But this intention only succeeded economically after the size and the price of geophones could be essentially reduced, and after shot hole drilling exhibited important advances. I remember we successfully used 5 geophones per trace early in 1940. But we had to stop this technique due to a lack of geophones. Here, too, we see that fundamental new knowledge alone is not decisive for progress, but very often technical-economic developments have to be added. In this case the development of new magnetic materials had been most important in reducing the size of geophones. A further advance for exploration seismics was the introduction of reproducible recording, which was made practicable by the development of magnetic recording during the middle and end of the fifties, although optical methods also have been used successfully. Here too the fundamental ideas were much older than the technical development. In this context mention the name of Frank Rieber, whose fundamental papers were from 1936 to 37. Due to reproducible recording a lot of computing processes could be carried outwith the data recorded.lt is interesting to learn that the proposals for these computing processes were on a rule older than the technique of magnetic recording and reproducing, in any case they were older than the digital processing of seismic data. I specially mention the time-domainfiltering, the optimal filtering, which had already been published in Geophysics Vol. XX, No. 2, by Simpson in 1955. I further mention deconvolution, which was already contained in a paper by Robinson in Geophysics Vol. XXII, No. 4, 1957; and furthermore static and dynamic corrections and in context with them the multiple coverage technique, which had already been invented by Harry Mayne in the early fifties. All these processes were intended to improve the recognizibility of the desired seismic signals and to enable a more precise determination of the times of these events, as weil as a better separation of such signals which have only a short time interval between each other. These processes are partly based on preceding mathematical papers or books. Here specially Norbert Wiener has to be mentioned. As to multiple coverage and the possibility of attenuating multiple reflections by this method, however, the origin of this idea seems to be independent of other sciences or techniques. As far as possible the computing processes mentioned were carried out by analog computers by 1960. Various instruments were designed to deliver corrected, filtered and stacked records. Magnetic delay-line-filters as described by Jones 1955 were built in order to filter in the time domain. By magnetic AM recording with mini heads it became possible to add seismic data. Etc. etc. . . By the commercial availability of Laser light sources it became possible to get very rapidly record,sections which had been treated with two-dimensional filtering in the frequency and wave-number domain. The most important breakthrough in the precise routine execution of all these desired computation processes was the introduction of digital computers in the mid sixties. This step had become possible by advances in electronics, especially by the development of transistors and integrated circuits. Moreover the invention of certain peripheral units, e. g. the convolver had been important for the practical use of computers in exploration seismics. The natural consequence of the use of digital computers for processing was the introduction of digital recording in the field in order to match the precision of recording to the precision of the digital processing. By solving this and other tasks the industry of manufactoring seismic instruments created the possibility of extracting continually more and more precise information from the seismic data. The Development of Energy Sources Shooting Technique The development of recording and processing the seismic data is certainly the most interesting and most important part of the history of exploration seismics up to now. But the development of the source of seismic energy must not be forgotten. After initial tests with weight dropping in 1910, Mintrop very soon decided to use dynamite as a seismic source. For a long time shots were detonated at the surface. The strong surface waves originating from this method could not impare refraction seismics. But the high consumption of explosives and disagreeable effects, as broken window panes etc. which often led to strong protest from the population, enforced shooting at a certain depth. This was the creation of shot hole drilling. Here applied seismies with its requirements has given a strong stimulus to the industry of shallow hole drilling and this stimulus then again had positive influence on exploration seismies. Very soon, i. e. as early as 1927, it was recognized that shooting below the near-surface low velocity layer, i. e. shooting below the weathered layer or below the ground water level, resulted in a most effective weakening of surface waves or ground roll. Besides central recording, shot hole drilling must be considered as being the second most important step in successfully introducing reflection seismies into practical use. At a later date jetting, percussion drilling or drilling with air circulation led to multiple shooting, specially to pattern shooting, in areas in which itwas difficult to get reliable reflections otherwise. For 10 or 15 years the seismic industry has aimed at developing further seismic sources other than dynamite. Perhaps the most interesting of the non-dynamite sources is included in the "Vibroseis" *-system, which has a certain analogy to the Chirp-method of Radar. John Crawford, the inventor of this famous method is also sitting among us. In marine seismies the design of non-dynamite sources has been especially intensified. Here the main effort has been in developing sources which are free of bubbles, or in which the disadvantages of bubbles have been cancelled. And here methods have been created which do not find any analogies in other sciences. * Trade Mark and Service Mark of Continental Oil Company
For refraction seismies the most general wave-front method was al ready applied in very early time, i. e. before 1930. But in reflection seismies, too, much has been done in this respect, whereby the concept of "migration " was created. Velocity as a function of the space coordinates was already considered approximately or precisely more than 20 to 30 years aga by the construction of curved rays and wave-fronts. A certain preparation of the data to simplify the construction of the reflecting interfaces was introduced by two kinds of corrections. By static corrections the influence of the weathering and the topography was allowed for and by normal-move-out-corrections, the shot point and the observation point were adjusted to a position half way between these two points so that the descending ray and the reflected ray coincided. These two corrections could al ready be carried out in the record sections more than 15 years aga by analog data centers as mentioned earlier. Moreover already at that time, e. g. with PRAKLA-SEISMOS, digital programs had been developed in order to infer the right position of the reflecting layers from times of the reflections picked in the records. Pitfalls in the interpretation of reflection records had also been recognized more than 20 years ago, as e. g. multiple reflections, diffractions and reflected refractions. When considering diffractions mainly Huygens principle is applied. Now, it is very interesting to learn that this principle is again essentially being applied in the most modern successes of data processing, i. e. in the production of migrated record sections. With this process which needs digital computers with specially large memories and computing velocities, the seismic reflection signals are transferred to that position in the subsurface, where they originate as reflections. A certain stimulus for the introduction of migrated record sections has been given by holography, though the analogy is not perfect. There is another way of describing the introduction of migrated record sections, namely we may say, that by this innovation the processes of picking reflections and of migration are reversed in order. That "picking of reflections", too, can be carried out automatically is a further step which has already been made but which has still to be improved. A further advance as compared to migrated record sections is the migration stack operation. With this process migration is executed before or together with stacking. We may say that exploration seismics, since the day of its birth, has reached a certain perfection as far as detection and localizing of inhomogeneities producing seismic signals are concerned. But as to the physicalchemical character of the inhomogeneities we are still at the very beginning. Relations between the reflection oscillations and the facies of the layers concerned or the gas-, oil- or water filling of their pores is still rather unknown. Here a bulk of work has still to be carried out. Outlook Ladies and gentlemen, I think the preceding review has shown how exploration seismics under the pressure of the ever increasing requirements of the oil industry and other clients has always endeavoured to look around in the domains of other sciences in order to find suggestions and ideas which might be suitable for incorporation into its own technique. But we must also point to the fact that individual ideas have been born within the industry which have no analogy in other techniques, e. g. multiple coverage. On the other hand certain ideas developed by exploration seismics have also been accepted by other sciences or techniques, e. g. pattern recording, which has been introduced in normal seismics quite recently under the name "array". Additionally I believe that exploration seismics has given a valuable stimulus to the digital computer industry. On the whole, I think, we may look back proudly upon the 50 years which have passed. But I also think, the achievements reached so far are an obligation for us to continue our researches and our designs and to check again and again the philosophy of our professional standing in order to comply with the many needs and problems which are still present. |