19	TREND SURFACE ANALYSIS-an exploitation technique worth considering!!! By STANLEY R. WHARTON

Introduction
It is no secret that the geological environment in Trinidad is complex, both structurally and stratigraphically, and that these complexities can be related both to the tectonics at the southern end of the Caribbean Plate and to the proto-Orinoco deltaic depositional environment. From an exploration and production standpoint, many have observed that we have reached a stage where the various techniques for finding oil in this complex environment have all been extended with varying successes, and that we are faced with several challenges which beckon us to employ innovative techniques to further improve our chances of finding oil.
We are also challenged in our efforts, to select the optimum combination of techniques that would provide the best results for precision decision-making. There is an urgent need for us to adopt these innovative techniques and incorporate new ideas in our exploitation concepts particularly since we have the facility of using workstations and sophisticated software programmes for analysing our data.
The question then arises as to what techniques to employ and how to use these techniques effectively.
Trend Surface Analysis Technique
Trend Surface Analysis is one of the oldest mathematical techniques that has been used consistently by geologists for the analysis of geological data and can be considered a possible candidate for exploitation of known reserves. The technique represents a methodology for manipulating map data and it produces a smooth approximation of a 3-D surface. The trend surface equation is a linear polynomial and the resultant trend surface is the graphical representation of the mathematical equation (Davis, 1986).
This mathematical method is used by geologists for separating map data into a regional component and one with local fluctuations. These two components are represented as (a) a Trend Map and (b) a Residual Map. The Trend Map can be used to define geometric models of 3-D shapes for basins or fault blocks, as planar, parabolic or cubic trends, depending on the selected polynomial equation for analysis.
The technique has been used successfully by petroleum explorationists in many parts of the world and its application includes the discovery of the famous Leduc Reefs in the Alberta Basin in Central Alberta, Canada.

Case Study - The Moruga East Field Trend Surface Analysis was carried out on a sub-set of data from the Moruga East Field using a 3rd Order Polynomial on the LANDMARK ZYCOR System, and both a Trend Map and Residual Map were generated from the raw data.
Using the Trend Map, the data was interpreted as showing a major anticlinal structure in the NW with less differentiated structure to the east and south (Fig. 1). The interpretation in these latter areas however, was constrained by the data set. The superimposition of a TLGM marker structure map on this Trend Map enabled an interpretation that suggest that the Field is represented by a faulted anticlinal structure with varying trap types depending on the fault types and the stratigraphy of the Field (Wharton, 1993 a, 1 993b).
In the analysis, the regional trend thus appears to be anticlinal with the greater part of the structure masked by the overlying younger Goudron sandstone. There appears to be a SW anticlinal trend from the adjacent double-nosed Galap Anticline area to the East but the younger Goudron sands does not allow that linkage of the structure to be seen on the surface. The anticlinal axes of both the ME main field and the double-nosed Galap Anticline have a displacement of approximately 1/2 mile in a right lateral sense, suggesting that tectonic activity was perhaps responsible for complex structuring in the area.
The Residual Map in comparison suggests that several 'anomalies' exist in the Field and these can be related to the complex geological structure (Fig.3) also. This data was integrated with interpreted data from both SAR photographs and aerial photographs, well log correlation and seismic data (Wharton 1993 a, 1993b), and the 'anomalies' were interpreted to represent the re-suits from a superimposed system of normal and thrust/reverse fault sets generated in the post Pliocene period. The positive and negative anomalies may be related to the relative elevation of upthrown and downthrown fault blocks that define the reservoirs in the Field. These anomalies when integrated with other data can be used to select prospective areas for drilling.

Hardware/Software Considerations vs. PC versus Workstation Platforms

In practice, there is a major difference in the computational capability of each of the platforms and this is based initially on the size of the numbers in the matrix of simultaneous equations for the test. This may be a function of the coordinate input data. Differences also exist in the output of the data as can be compared from an analysis that was performed on a PC (Fig.4) and another on a workstation (Fig. 1). Experience shows that there is a major difference in the output from the Trend Analysis derived from the small PC package (COGS), from the analyses derived from the more sophisticated and flexible Workstation package (LANDMARK Zycor). In analysing the results a major difference was identified in the output of both the Trend Map (Fig. 1 versus Fig.4) and the Residual Map for a similar data set extracted from the Field. A comparison between the functionality of the software on the two platforms is presented in Table 1.

Challenges in the Future
One of the challenges facing geologists in the petroleum industry in light of active exploration is whether the Trend Surface Analysis technique and other similar techniques have any credibility their area of interest. The other challenge relates to the effective utilisation of the sophisticated hardware and software programs which are available to the geologists, to produce optimum results for decision-making. Geologists would have to depart from two familiar syndromes: the "Default Syndrome", where the computational procedures on the computer for any given software program are pre-set and all that is necessary is a 'push of the button', and the "Black box Syndrome", where the computer software programs are thought of as providing the 'panacea' for all geological problems.
In reality, we will have to become more interactive with the software programs so as to select the correct sequence of algorithms that best allows for the most efficient working of the data and provide the best interpretation within the available constraints. And this can be achieved only by a thorough analysis of the problems that are presented to us and a hard thought of the solutions that are required to provide the desired result. Thus, geologists would have to develop an understanding of the analytical techniques that the software programs offer in order to select the correct parameters.
One common problem relates to which algorithm should be used as opposed to another especially for programs which provide the flexibility of choice. The selection of a proper sequence of algorithms however invariably comes with experience and practice, which can only be developed from repeated trials and comparisons, and checks made on the validity of the results. This may be tedious but is a necessary process.
Geologists should also be aware of computer programs that do not work the data as efficiently as some of the larger packages and should examine the algorithms used and compare the results. A case in point is the comparison of the results from Trend Surface Analysis on a subset of data from the Moruga East Field using a small PC package and a package on a Workstation.
As to the use of the Trend Surface Analysis technique in our environment, I believe that the technique has credibility as can be seen from its use in the Moruga East Field. In order to know the real benefits, however, we would have to actually per-form the analyses and compare the results with other data. Only then will we be able to make an objective criticism and suggest alternative approaches. The Trend Surface Analysis test has a proven track record internationally, since the l950's, and its use has enabled the discovery of the large oil and gas fields in the Alberta Basin, Alberta, Canada.
While not advocating that Trend Surface Analysis solely would assist us in finding large quantities of hydrocarbons, its use with the integration of other data can assist in understanding the structural characteristics of our reservoirs which could further improve our leads for better decision-making.

References
Davis, J.C. 1986. Statistics and Data Analysis in Geology. John Wiley and Sons.

Wharton, S.R. 1993 a. Exploration Within A Complex Thrust-fold System - Moruga East Field,
Trinidad. AAPG/SVG International Congress and Exhibition. Caracas, Venezuela. March 1993.

Wharton, S.R. 1993 b. An Integrated Approach to the Study of Pliocene Gros Morne Reservoir Sands For Developing Play Concepts - Moruga East Field, Trinidad. Society of Petroleum Engineers, 11th Technical Conference and Exhibition. Trinidad. June 23-25, 1993, (In press).

Home | About Us | Links To The World  |  Publications | Upcoming Events |

Feedback | Executive Members | Geology |Teaching Resources| Virtual Field Trip

THE GEOLOGICAL SOCIETY OF TRINIDAD & TOBAGO
P.O. Box 3524, La Romain, Republic of Trinidad and Tobago, W.I.

Comments or Questions?