An Introduction for Petroleum Engineers
Richard Duncan & Allen Satterwhite
Please visit our Web Page LSU PETE Dept.
Table of Contents
Seismic Internet Resource Sites
Reflection seismology uses artificially generated sound waves to obtain subsurface geological information for engineering applications. Today, virtually all oil and gas companies rely on seismic interpretation to choose locations for exploratory drilling. Futhermore, with the advent of technological advances in the areas data processing, seismology is now being expanded to include production and reservoir management strategies. Although in some specific situations the presence of hydrocarbons may be inferred from seismic interpretation, it is important to remember that exploration seismology does not directly map the presence of oil and gas. Instead, seismic interpretation leads to the mapping of geologic structures which may contain the oil and gas.
Reflection seismology consists of measuring the time required for a sound wave to travel from a source located at the surface, to a subsurface structure, and back to the surface to a receiver SPREAD. At each geological interface (ie, sandstone/shale interface), a portion of the sound wave will be reflected while the rest of the energy is refracted or transmitted deeper into the earth until the next interface is encountered, upon which a portion of the wave will once again be reflected and refracted. The reflected data recorded by each receiver is known as a trace. A trace consists of plotting the amplitude of the reflected wave vs time NMO . Traces over a given area can then be analyzed as cross sections to identify geological trends. The end result is a pattern of black and white lines vs time or depth SEIS2. . It must be remembered that this seismic record is a recording of a huge number of reflected acoustic waves that have been processed, re-processed, converted, inverted and manipulated to hopefully more accurately reflect the underlying geology. Although the resulting plots can be used to infer the presence of geologic structures; and in limited cases the presence of hydrocarbons, the ultimate exploration tool remains the drill bit.
In 1898, Milne first suggested that seismic waves might be used to describe subsurface conditions. During World War I, both the Germans and the Allies investigated the use of seismographs to discover the location of 'enemy' artillery. In 1926, a German patent was issued to Ludger Mintrop which noted that the 'seismometer' would work better to determine rock structures than a divining rod! Seismic exploration for oil began soon after. In 1924, oil was discovered both in Texas and in Mexico at locations chosen by seismic methods. However, these early techniques mainly used refraction, rather than reflection method. By 1930, seismic exploration had been used on five continents.
Below is a reprint of an old seismic reflection survey from California. Note that there are 11 tracks indicating the use of a 10-channel recorder. These are 10 different recordings placed side-by-side. They were originally recorded on a drum and were presented sideways rather than vertically SEIS1 . Modern surveys are shown vertically with one half of the curve colored for ease of interpretation. Compare the older chart from above with the modern chart below SEIS2 .
Sesimology is the branch of geophysics which studies the movement of seismic waves through the earth. A seismic wave is a sound vibration or perturbation which travels through the earth. While the term seismology is generally associated with naturally occurring seismic events such as those produced by earthquakes, reflection seismology specifically deals with man made or induced vibrations in the earth. The goal of reflection seismology is to interpret subsurface features by analyzing the response of sound waves which have been artificially generated and reflected back to a recorder at the surface. The recorded data provides valuable information relating the velocity and travel time of the sound waves to the depths of corresponding subsurface features.
The generation of sound waves is provided by a source. The most commonly used land sources include explosives, and vibrating trucks. Other types of sources which have been used include weight-dropping units and land air guns. The choice of which type of land source to use depends upon criteria such as the desired depth of investigation, accessibility of location, repeatability, and safety factors. For example, in remote wilderness or mountainous areas where road access is severely limited, dynamite may be the most practical source. However, near commercial or residential areas the use of vibrating trucks would be much more practical. The use of explosives as sources requires drilling a small hole in the ground, placing the charge into the hole, and filling the hole back up to tamp the explosion and direct the energy down into the ground. Vibrator trucks use hydraulic pistons to oscillate a mass in a vertical manner which imparts a force through an attached base plate into the ground. The transmitted force in turn induces vibrations down into the earth .
In land surveys the reflected sound waves are recorded using geophones. Modern geophones utilize moving-coil electromagnetic technology. A coil of finely wrapped wire is suspended inside a circular slot cut into a cylindrical magnet. The geophones are placed in firm contact with the ground in a spread (layout) which will be discussed later. As the ground moves or oscillates due to the reflected sound waves, the magnet will move but the coil remains fixed. A voltage is then generated due to the relative motion between the coil and magnet, which in turn is associated with a given amplitude of the reflected wave.
In a marine environment, pneumatic air guns serve as the source for reflection seismology. Several air guns may be used in conjunction in a designated firing pattern as they are pulled near the back of the boat. They generate sound waves by releasing a burst of extremely high-pressured air into the water. Reflections may then be recorded using one of two techniques. First, seismic streamers or cables may be pulled behind the vessel boat schematic . These consist of a series of long buoyant cables which contain the recorders, called hydrophones. As the vessel moves, both the air guns and streamers are towed behind and constantly survey the area below streamer in tow . Another type of marine receiver is the ocean bottom cable (OBC). This requires laying the cable containing the recorders down on the sea floor instead of towing them behind the boat. This approach is typically utilized in shallower, high traffic areas where pulling long streamers behind a vessel is not practical. In this scenario the vessel would still move about the area generating sound waves in the water, but the reflections would be recorded on the ocean floor by the OBC.
In order to properly place recorded data, relative and absolute positioning of sources and recorders is obtained utilizing sophisticated positioning techniques such as electronic distance measuring, global positioning systems, radio positioning, and satellite observations. Perhaps the most significant technological advancement in conducting seismic surveys has been in the area of data processing. With the evolution of faster more powerful computers, greater volumes of data can be processed in a given time. As a result, a more thorough and descriptive analysis of the subsurface can be performed with less interpretation error.
Although both body waves (P-waves) and shear waves (S-waves) are present, normally only P-waves are used in reflection seismology. P-waves are compressional waves which travel through a medium by axial deformation of the subject material. S-waves are like waves in a guitar string that travel up-and-down or side-to-side waves . P-waves travel considerably faster than S-waves (Vs = 0.7 Vp), and as a result will return to the receiver first. While both liquid and solid materials will transmit P-waves, only solid materials will support S-waves . All of the wave characteristics discussed from here on will be that of P-waves unless otherwise stated.
Sound waves travel through the earth in a spherical manner. A wave front represents all points in a sound wave that have traveled for the same amount of time, but not necessarily the same distance. Rays are lines that describe the direction of propagation of waves and are at right angles to the wave front. In reflection seismology, calculations are simplified by utilizing mathematical assumptions which allow us to substitute rays for spherical wave fronts. When sound waves are generated, the rays representing the direction of propagation will travel in straight lines until they intersect an interface. When the ray strikes an interface it is partitioned, or split into four separate waves. A P-wave will generate reflected and refracted P-waves and reflected and refracted S-waves. Similarly S-waves will also partition into reflected and refracted waves. Reflected waves represent the portion of energy which bounces off of the interface and returns to the surface. Refracted waves represent the portion of energy which has been transmitted through the interface deeper into the earth. Upon encountering the next interface, a portion of the energy is once again reflected and refracted. Refracted waves also describes those waves which upon contact with an interface, travel horizontally along the interface and eventually reflect back up to the surface some distance later. This special type of refracted wave, known as head-waves, are used in refraction seismology. However, due to certain limitations associated with refraction seismology, most modern exploration seismology makes use of reflected waves, hence the name reflection seismology.
The amount of energy that is reflected from upon contact with an interface depends on the contrast in impedance at each interface. Impedance is the product of the velocity and density of each layer. Since density and velocity contrasts are normally quite small between two layers, the amount of energy reflected is typically quite small, usually much less than 1%. Therefore, most of the energy is refracted or passed through the interface into the next layer. However, the wave does not continue in a straight line. Instead it is refracted or bent at the interface. The velocity at which sound will travel through each layer is dependent upon characteristics of the rock matrix and interstitial water.
Determining the degree of refraction is quite simple mathematically. As shown in the drawing below, an incident wave is passing through Layer 1 with a velocity of 2500 m/sec when it impinges on the interface at angle A1. The reflected part of the wave returns through Layer 1 at the same velocity and at the angle of incidence. The refracted wave, however, is bent to angle A2 according to Snell's Law as it speeds up to its new velocity of 2800 m/sec.
(Snell's Law)
As can be seen in the equation and in the drawing, when a wave enters a layer with a higher velocity, the refraction angle (A2) will be greater than the incidence angle (A1). This has the effect of bending the wave back towards the surface. Since velocity normally increases with depth, as the wave moves deeper into the earth, the refracted wave will become more parallel to an interface.
Head waves, after traveling some horizontal distance along an interface, can and do return to the surface at some point. Huygen's principle states, in brief, that any point along a wave can be considered to be the starting point of a new wave. Therefore, in the drawing below, wave A (ignoring the reflected wave for the moment) is refracted along the interface, leaves the interface at the same angle that it entered, and returns to a receiver. The refracted wave travels along the interface at the velocity of second layer and therefore is traveling faster than either the incident wave or the returning wave.
Because the refracted wave had a higher velocity than a reflected wave for part of its length, at some point the refracted wave will return to the surface before the reflected wave. This point is called the critical distance. The equation for this is shown below:
Since in reflection seismology normally only reflected waves are used, the geophysicist needs to determine the horizontal distance from the source to the recorder at which point the head wave for a given interface will arrive before the reflected wave. Normally the receivers extend no more than twice the depth of the first reflective layer.
The method or layout used to obtain seismic data varies depending upon factors such as geologic objectives, accessibility of area, land or marine environment, and surface obstructions. The spread of a given field layout describes the relative locations of sources to the recorders. However, one commonality is that the recorders are placed in linear fashion with respect to each other to insure that changes in reflected data between recorders is attributable to geologic factors rather than changing field conditions or gathering techniques layout . As sound is reflected from interfaces and recorded at evenly spaced receivers, the two way travel time (TWTT) for the reflections are plotted versus offset (distance from the source to the receiver of interest ). In other words, the trace for all of the receivers are lined up (in respect to the corresponding offset) and plotted versus TWTT. Trends representing reflecting interfaces will then be identified as hyperbolas extending across the traces. This distribution is due to the mathematical relationship relating velocity and offset distance. This geometric effect is then corrected using a procedure know as normal moveout (NMO). If when NMO has been accounted for the reflecting surfaces across the traces have not been corrected back to the horizontal position, then it is possible the reflective layer is dipping NMO .
Dip moveout, or DMO then attempts to correct the seismic recording for the remaining deviation from the horizontal. If this procedure corrects for the difference, than the degree of dip can be calculated as the slope of the deviation DMO2. . The first seismic recording in the drawing above shows a recording of a dipping bed. The second recording shows a recording of a flat interface with a NMO of 13 ms over a 400 m geophone spread. If the bed was not dipping, the NMO correction would correct the hyperbola to a straight line, which would accurately reflect the nature of the bedding. The third recording above shows the resulting dip by subtracting the NMO from the original recording. If anomalies remain in the distribution of data, other factors such as migration may account for this problem. Migration effects will be discussed briefly later in this paper.
The predominant field method for obtaining seismic data today is known as the common-midpoint method (CMP). This technique involves moving the location of the source in respect to the recorders such that a given subsurface point is sampled more than once and recorded by different recorders. Remember that a trace is a recording of reflected signals for each receiver. By sampling a given point with different receivers, several traces can then be stacked or combined to remove some of the basic sources of error such as timing errors or background noise. In simplest terms, we have now sampled a given point multiple times to reduce interpretation error for each reflection. Once the data has been recorded using CMP, corrected for NMO and DMO, and stacked, anomalies may still remain in the data. When performing the above procedures, we assumed an average velocity of the depth of investigation. For the most part this assumption introduces only small amounts of error which can be negated. However, in some situations geologic conditions effect velocity profiles in such a manner as to introduce significant error. Migration attempts to account for these differences by varying the velocity profile which has been used to interpret the data. This process may be performed either before of after stacking, hence the terms pre and post stack migration. After these operations have been performed, we may represent the reflections in a plot of offset traces vs TWTT or converted depth to analyze the corresponding subsurface interfaces. It should be noted that the description of the above process to convert raw recorded data to a final geologic cross section has been greatly simplified here. These represent only the major categories under which many much more complicated calculations and procedures are performed.
In 2-D seismic surveys, data are collected along lines forming a grid of sorts. 3-D models of the subsurface are then constructed by interpolating between the 2-D seismic lines. Unfortunately, features appearing on the 2-D lines may actually be offset and small features such as splinter faults may fall in-between lines and go undetected. As a result, significant interpretation error can occur in geologically complex areas. 3-D seismic attempts to eliminate this type of error by shooting data points uniformly over an area rather than simply along lines. In simplest terms, 3-D seismic obtains much more data over the same area. However, 3-D seismic requires much more computing power and processing time than 2-D seismic. 3-D seismic has only been recently made feasible by the advent of faster more powerful computers. Using sophisticated software packages and graphic interfaces, we can produce advanced 3-D images of the subsurface. Today, industry is studying the implementation of 4-D seismic. This involves implementing time as the fourth dimension. Reservoir properties such as fluid saturations, temperatures, and fluid interfaces can be monitored with 3-D seismic to account for changes in the reservoir with time. This process is allowing for more effective reservoir and production management strategies to maximize production revenues.
ADDITIONAL SEISMIC INTERNET RESOURCE SITES
Input/Output Seismic Processors
This page last updated on 11/4/97
