ChemTerra International

GEL in-depth
Surface Exploration Geochemistry (SGE):
The Development and Description of the GEL System


Introduction

Surface exploration methods, including soil gas surveys, have been proposed for the oil industry over the past six decades. At ChemTerra International (CTI), we are introducing a new approach for instant and integrative soil gas surveys by solving essential problems inherent to the general concept of micro-seep surface exploration.

Rather than following the often ill-defined arguments of proponents and opponents of these methods, we at CTI focused on the key question as to what extent varies SGE (Surface Geochemical Exploration) methods could contribute to exploration success. After a decade of R&D testing and investigating a large number of proposed SGE methods in the late 1980's and early 1990's (involving a $8 million effort), CTI developed and introduced the GEL (Geochemical Exploration Lead) system, a high profile gas sampling / compound analysis / numerical analysis system of extreme, quantifiable sensitivity that uses a seep fingerprint approach as a defined seep signal in fuzzy and noisy soil gas data. Key emphasis in this particular SGE approach were the ease of day-to-day use, the universal application in almost any environment, and a long field / application test period under the guidance of an exploration department with a solid prospect track record (GEL Track Record).

The GEL system evolved from the former Gulf Oil surface gas exploration technique described in detail by Jones and Drozd (1983) and subsequent extensive R&D in-house efforts to improve sampling, analytical data acquisition, noise reduction, seep recognition, and seep rate calculations to a level unparalleled by any other SGE technique.

Soil Gas Magnitude Data and Their Significance for Micro-Seeps

Conventional soil gas surveys are among the most common SGE technique applied. These gas surveys typically record the magnitude of an array of gases, including C1-C4 alkane gases, which occupy part of the soil porosity. The sampling techniques involve the instant magnitude of free gas volumes in the soil/sediment, or gas collection over defined time period intervals as an integral signal. The concept of these soil gas surveys is based on the frequent observation of visual (macro-) gas and oil seeps in association with hydrocarbon (HC) reserves at depth. Link (1952) points out the many substantial field discoveries made during the past decades that are associated with surface hydrocarbon (HC) seeps.

This concept of macro-seeps as an indicator for deeper HC reserves can be extrapolated to the micro-seep level with the argument, that many, if not most, oil and gas fields leak some HC gases to the surface which are too subtle for immediate field recognition, but can be detected by modern, highly sensitive analytic techniques. Thus, increased HC soil gas readings - in particular the C1-C4 HC fraction as the most mobile, fugative reservoir HC's - in a given work area are usually considered to be indicative for potential reserves at depth.

While this general concept of seeps as direct surface indicators is correct in principal, it is often ignored that - in contrast to visible macro-seeps - the signal in seeps on a micro-level is small and often misleading for two reasons:
  1. First, the S/N ratio of micro-seeps is magnitudes smaller compared to macro-seeps. Surface gas "anomalies" (as well as "anomalies"of other SGE techniques) are often defined from data statistics, but with considerable uncertainty as to the significance of these anomalous surface features.
  2. Secondly, the first discovery of biogenic alkanes other than methane by Davies and Squires in 1954 and subsequently the report of C2+ alkanes in soils and roots combined with remarks on the limited value of surface prospecting by Smith and Ellis in 1963 cast a shadow on the application and value of soil gas magnitude readings in particular and surface exploration in general.
The AAPG Bulletin article of Smith and Ellis (1963) was heavily disputed by senior surface explorationists such as Horvitz (1972) and Davidson (1994). However, since the 1970's the geo- and biochemical scientific community has assembled a massive body of evidence for the occurrence of low level C2-C4 HC gases of biogenic and low-temperature (non-reservoir) origin in soil and shallow aquatic environments. The occurrence of these soil in-situ HC gases is highly variable and erratic due to a number of factors controlling the formation, the selective degradation, dissolution and diffusion of typical soil gases. Thus, it is not surprising that most of these surface gas readings have a fuzzy appearance with the possibility of high surface gas readings and clusters being false positive anomalies (e.g. Weissenburger, 1991, 1996) with no relation to subsurface reservoir HC.

CTI in-house gas data fully confirm the complex nature of surface HC gases and the difficulty of data interpretation. Surface C2-C4 HC gases in highly variable amounts are encountered in both barren areas such as the Canadian Shield and in oil-bearing country. Consequently, HC soil gas magnitude anomalies do not necessarily reflect true micro-seeps. This observation also has considerable bearing on "indirect methods", that are believed to reflect presence or absence of HC gases. The key lesson here is that "presence" or "magnitude" of soil HC does not equate to "micro-seep" as often presumed.

However, many observations - including CTI's data collection of about 100,000 samples taken worldwide clearly show that soil gas magnitude data taken from shallow drill holes have some tendency to cluster over producing fields and successful prospects. However, a one-to-one correlation does not exist and cannot be expected for reasons discussed above; the diagnostic value of many of these surveys - instant or integrative - displaying soil gas concentration is, therefore, rather limited. This judgment is also in agreement with the more critical literature by Hunt (1979, 1981, 1995), Schoell (1984), Price (1986), and Hunt et al. (2002). This judgment can also be passed on to essentially all "pattern" and "indirect" methods relying directly or indirectly on gas magnitude data: high surface gas readings may be triggered by seeping HC as well as soil in-situ HC. For instance, a reducing soil environment with high HC gas concentrations may not be associated with active or passive seepage; it may reflect regional or local specific soil / sediment conditions unrelated to seepage from reservoired HC at depth.

This body of questions and problems led CTI to the question: "How can SGE technology be improved to the level of a reliable technique with confident results useful for exploration and applicable in day-to-day operation in almost any environment?"

ChemTerra International's (CTI) Concept

Given the fact of some underlying exploration "value" is provided from basic, standard soil gas surveys commonly applied in the industry, CTI started a rigorous R&D program - heavily supported by the Canadian oil industry. The key focus of this program was the investigation of the following questions:
  • Do micro-seep HC chimneys exist in the subsurface over oil and gas fields?
  • To what quantifiable extent do simple soil HC magnitude readings relate to deposits in the subsurface?
  • What is the nature and source of background and noise gases?
  • Can noise gases be recognized and reduced to enhance the S/N ratio of the micro-seep signal?
  • Can a true HC seep signal be retrieved from bulk noise and varying background surface gases?
  • If detected with confidence, are micro-seep HC gases in the C1-C4 range the result of preferably vertical gas migration from depth? Vertical gas migration is the common presumption of all surface detection methods.
  • To what extent are genuine C1-C4 HC micro-seeps related to existing and new production underneath?
The overall goal of this R&D program was to develop better (or new) procedures that:
  • Considerably improve the information value of soil gas surveys for exploration.
  • Clearly increase exploration opportunities and decrease drilling risk.
  • Provide high quality information value for exploration at reasonable cost.
Top of the page

CTI's Research

CTI's investigations in the late 1980's on log records, rock properties, and rock chemistry confirmed earlier literature reports and claims on the presence of rock alterations in HC chimneys induced from seeping reservoir HC gases. At present time there is an overwhelming literature body available of the direct and indirect evidence for HC chimneys in onshore and offshore environments. The Ekofisk (North Sea) HC chimney, Figure 1, recognized as a seismic wipe-out zone as a result of seep HC (van den Bark and Thomas, 1980), was a starting point for subsequent research and now serves as a textbook example for HC chimneys.

Figure 1


Today, more refined seismic data processing routines identify characteristic HC seep patterns in chimneys as shown in Figure 2 (Aminzadeh and Connolly, 2002):

Figure 2


The recognition of these HC chimneys and their link to the surface in form of HC and rock alteration anomalies is important information for two critical reasons:
  1. There is widespread proof for the existence of HC chimneys on a macro- and micro-scale.
  2. There is proof for a preferred vertical migration of these seep gases.
All Surface Geochemical Exploration (SGE) techniques rely on vertical migration. Although not an absolute rule, the migration of gases has a strong tendency for vertical ascension - in contrast to liquid HC. The Western Canadian Basin is a textbook example of lateral oil migration, - but with dominantly vertical gas migration. This observation is one of several reasons for the GEL system to exclude the liquid HC (molecular mass range > 60) from analysis.

Thus, the search for genuine surface seep gas anomalies or (indirect) surface alterations would make logical sense in the effort to support oil and gas exploration as a new, reliable tool independent from subsurface data and models.
Top of the page

CTI's Development: The GEL Seep-Fingerprint Approach

The problem of gases of different origins co-mingling in the surface environment can be numerically solved. The surface gas technology, termed GEL (Geochemical Exploration Lead), was developed at CTI to do exactly this. GEL applies linear vector-based techniques to high precision analytical gas data sets to numerically test for, isolate and reconstruct true seep gas signals and separate those from noise and bulk surface gases of different (non-migration) origin. This seep-fingerprint is numerically expressed and reconstructed in form of a seep gas family present (or absent in case of non-seep environments!) among non-seep gas families. In essence, if active seepage is occurring, this seepage leaves a typical seep-fingerprint in field samples located in a seep site. This seep-fingerprint signal varies according to seep type and intensity, with a possible range from 0% (no seep contribution) to 100% (absence of non-seep HC gases). In the absence of active seepage, no such seep-fingerprint pattern evolves from numerical data processing. This numerical processing step is essential as it provides a critical filter to eliminate noise and false signal data that go undetected in conventional surveys which simply record or time-accumulate gas magnitudes. In its principal, the GEL method follows the routine evaluation of seismic data noise reduction by applying appropriate (numerical) filter techniques.

The ability to detect and reconstruct a defined seep signal is a corner stone of the GEL procedure: Numerical failure to reconstruct such a seep gas fingerprint from surface gas data is clearly indicative of absence of seepage, irrespective of any presence or magnitude of soil HC gases. Absence of active seepage, in turn, relates to increased drilling risk for prospects.

The key approach of the GEL system is based on sound principals with a clear concept to identify micro-seepage and to lower the uncertainty associated with "anomalies" of other techniques:
  • The GEL system utilizes deep gas sampling from about 1.00 - 1.20m depth - away from the zone of maximum biological soil activity at 10-30cm depth with biological C2+ HC noise gases in production from the decay and transformation of organic debris in this soil zone of high total organic carbon (TOC). Data of ZoBell (1944) illustrate the exponential decline of biological soil activity with soil depth. More recently, Klusman (2003) presented seep data showing the exponential decline of seep gases as they approach the near surface zone. In consequence of these adverse relations of exponentially increasing biological noise gases with exponentially decreasing seep signal towards the surface the very shallow soil zone is the least favourable sampling location. Smith and Ellis (1963), and later Weissenberger (1991), point out the pitfalls of surface exploration sampling in soils of variable TOC and biological activity. GEL sampling focuses on the economic depth interval where seeping HC content is maximized and biological/residual noise is sharply reduced.


  • Deep GEL sampling also eliminates or largely reduces other surface effects affecting sampling in the very near surface range: deep gas samples are from an environment of relatively stable moisture, temperature etc. Typical atmospheric disturbances and extremes are minimized at GEL sampling depths. Equally important, pervasive surface HC contamination – in particular over oil fields – is eliminated from deep gas sampling. As a result, instantaneous sampling is sufficient and far advanced from so-called "time-integrative" sampling at dubiously shallow depth of 10-30cm.


  • The GEL sampling involves a gas volume to maximize concentration of the HC soil gas.


  • The GEL system recognizes that C1-C4 HC soil gases are pervasive and ubiquitous, irrespective of subsurface conditions. In fact, the bulk HC surface gases may be of non-reservoir origin - despite the deep GEL gas sampling.
  • The GEL system focuses on the molecular mass range from < 60, the essential mass range of practically all migrating and seeping HC gases (see e.g., Price, 1986; Hunt, 1995)


  • The GEL system deliberately avoids the higher molecular mass range > 60, because this mass range contains only traces of migrating reservoir components. Incorporating this higher mass range into data sets adds noise and confuses the issue.


  • If active seepage is occurring in an area, samples taken in that area have a portion of their total C1-C4 HC gases originating from this seepage. The seepage leaves a defined gas-chemical fingerprint in these samples as dictated from the reservoir content and reservoir conditions, and the widely accepted seep mechanism of reservoir gas micro-bubbles escaping along a micro-fracture network. The GEL seep-fingerprint signal does not depend on "learning data sets" because the surface gas response from a HC-filled, leaking reservoir is defined.


  • The GEL seep-fingerprint can be recognized and quantified from numerical data processing using vector algorithms and linear equations to numerically extract this seep fingerprint – if a HC seep contribution is present in some field samples.


  • Because the GEL procedure is built around numerically tracking seep gases in surface gas samples as a response of reservoir leakage, and since no learning data set is required, the enormous potential of oil field contamination being reflected in "learning data sets" is eliminated. In fact, many of the early "classic" anomalies in SGE (such as Cement Field) are now proven to originate from oil field contamination (Reynolds et al, 1990; 1991). However, existing fields can be carefully used to test GEL as an exploration tool or to test the area for the capability of seepage. Results are reliable because of the deep GEL sampling away from surface contamination.


  • The presence or absence of the seep-fingerprint is a distinct and unique geochemical indication of exploration / prospect viability.
Top of the page

OilGEL and GasGEL Values and their Track Record

The question arises how GEL actually measures or expresses seepage from soil gas data taken in the field. With respect to the GEL system design, three seep criteria are significant:
  1. Active C1 (C1 - C4) seep gases add to universally occurring soil in-situ biogenic hydrocarbon gases. Areas of seepage are, therefore, often areas of elevated total hydrocarbon gas concentrations above a regional background.


  2. If active HC seepage is present in an area, the seeping HC leave a (defined) "seep fingerprint" in these gas samples. The GEL procedure reconstructs this seep-fingerprint signal from numerical data exploration in form of seep gas family.


  3. Reservoir gases are an independent source or "Family" of gases. This seep fingerprint, derived from reservoirs, and gases from other sources mix in shallow soil environments. Seep environments are, therefore, characterized by soil gas samples containing a proportion of the seep fingerprint gases derived from the seepage process.
These criteria above form the basis of the GasGEL and OilGEL factors as numerical expressions of active seepage, either dominantly dry gas seepage (GasGEL) or wet gas seepage (OilGEL). The heart of the GEL system is these calculated GasGEL and OilGEL values. GasGEL and OilGEL values are the result of a systematic, logical and consequential deterministic approach to recognize and quantify HC seepage. GasGEL and/or OilGEL values are mapped and highlighted in CTI's extensive reports. It is a common scientific principle of CTI's reporting to list both raw and processed data.

So far, more than 80 prospects have been GEL surveyed, interpreted, and tested. The overall correct prediction using an integrated (geology, seismic and GEL approach relates to increased exploration success by a factor of 1.5 to 2.0, depending on area; note, that this record is achieved on a prospect-by-prospect basis, not on a well-by-well count (GEL Track Record).

Integrative GEL System: iGEL

Usually a single (instantaneous) GEL survey is sufficient to provide superior exploration information on the HC charge of an area or over prospects. The retention/residence time of seep gases in the soil is usually sufficient to record a good seep signal from a single, one-time GEL survey because of the unsurpassed GEL sensitivity to seepage (see below) and the deep field sampling at depth levels undisturbed from human/atmospheric/vegetation disturbances. However, the data and signal quality can be further enhanced by using the integrative GEL system (iGEL). This may be critical and of advantage for prime exploration targets in very low seepage environments where geologic conditions (e.g. good seals) minimize the seep rate. iGEL involves stationary field probes for multiple gas sampling at defined time intervals. The result is an impressive S/N enhancement from stacking of GEL data with a priori high S/N ratio. iGEL may involve two or more GEL sampling phases. The additional cost of iGEL over GEL are low in most survey applications because the repeat sampling is fast and along tracks and lines established from initial GEL sampling.

GEL Sensitivity from Seepage Simulation Studies

Because GEL involves the numerical reconstruction of a HC seep signal from soil gas data, the GEL response to seep rate changes can be easily quantified and visualized from seep simulation studies. The GEL procedure is the only advanced direct SGE technique that can demonstrate the direct increase in "GEL values" from increased seep rates.

Seep simulation studies show that the OilGEL value increases by a factor of 4 when the HC seep magnitude doubles. This extreme sensitivity to seepage is achieved, because the GEL system is deterministic and focused on seep identification among bulk noise gases. No statistical or "pattern"-based system can achieve this sensitivity level in regard to seepage monitoring. In fact, none of the SGE methods are in a position to relate their signal strength directly to seepage intensities.
Top of the page

GEL Reproducibility

Instant soil gas data usually show some or even large variability in repeat surveys. This variability and the assumed low reproducibility have been used in the past to discredit the application of such surveys, but with no or little investigation on the causes of this variability. Several factors cause the variability:
  1. Active seepage is a dynamic process and some variability of the seep magnitude is expected. Figure 2 clearly illustrates the dynamic character of HC gases migrating along a chimney.


  2. Most of the variability of soil HC gases is, however, not associated with true HC seep gases, but with in-situ biological noise gases, often forming the bulk HC portion in the soil matrix. Since biological soil activity is highly variable depending on changing conditions, the variability of noise gases is not surprising.


  3. Multiple GEL repeat surveys in virgin areas and over established fields show consistent results, although the dynamic character of the seepage is recognized.
The Day-to-Day Operation of GEL and GEL Reporting

CTI carries out its own field and lab operation. CTI owns a fleet of ATV's to operate in all terrains during all seasons. Field sampling is usually carried out by CTI personnel to ensure highest sampling quality.

GEL samples are taken in a grid formation or along selected (seismic) traverses and lines. Strict sampling guidelines are provided for sampling density and off-on locations. Since seepage is also controlled by the fracture network (no "blanket emanations"!) the GEL system requires a certain sampling density as a function of prospect size and the mode of operation (detailed survey versus reconnaissance survey). The argument of few samples being sufficient is refuted for any SGE tool because at least some field samples have to be close or over micro-fractures where possible seepage occurs. The lower the sampling density, the higher the chance of sampling between fracture locations with no HC seep contribution.

Lab analysis is carried out in-house on instrumentation exclusively dedicated for soil gas analysis with maximum sensitivity/performance, complete chromatography baseline separation of all gas constituents, and unambiguous compound identification/quantification. Common lab cross-contamination from various types of samples is, therefore, impossible at CTI's labs.

Once the raw analytical data are quality controlled, a data matrix is formed for numerical processing to "extract" a potential seep-fingerprint from data sets. In case true micro-seepage is observed, the micro-seepage signal is evaluated into Class A-C seepage, depending on the quality and intensity of the micro-seepage.

All reports are extensive volumes of documentation of a) raw data, b) processed data, c) description of the GEL procedure, d) description and interpretation of results, e) essential data plotted on topographic maps f) summary and conclusions, g) executive summary, and h) all data listings and the report both printed and in convenient CD digital format. Depending on contract, the client exclusively owns the data or shares the data with CTI.

In essence, CTI provides a complete turnkey operation from survey planning to the final stage of report presentation and discussion at the client's office or headquarter.

Benefits from GEL Surveys

Clients of CTI are presently using the GEL system on an operational basis to:
  • evaluate geologic prospects
  • prescreen large areas, allowing for a very focused seismic/geological evaluation
  • cost effectively evaluate exploration potential in areas of environmental concern
  • evaluate stratigraphic traps that are seismically difficult to identify
The benefits from GEL surveys are the following:
  • GEL surveys identify micro-seep areas, which in turn directly indicate the presence of subsurface hydrocarbons underneath (gases preferably migrate vertically up in the section).
  • Inexpensive GEL surveys provide guidance for substantially more expensive seismic surveys.
  • GEL seep results are exploration data independent from subsurface models or geologic concepts. They verify or dispute geologic concepts.
  • GEL data are indicators for the HC-charge or non-charge of a prospect.
    •
It is stressed at this point that the GEL procedure outlined here is not a panacea, but a useful auxiliary exploration tool independent of any seismic or geologic concept. With this in mind:

the essential aspect of GEL surveys is the reduction of drilling risk (GEL Track Record)

The significant result is a cost reduction of up to 50% per discovered oil/gas unit when GEL Data are integrated into conventional exploration decisions.
Top of the page

Regional GEL Survey Examples

Two regional examples may illustrate the application in the field in day-to-day operation:

Figure 3
Figure 3 shows a long regional survey line over the Brassey oil field in northeast British Columbia, Canada. The line graph represents the total C2-C4 hydrocarbon concentration in ppm. By applying the GEL numerical filter technique to the total data set, surface gas samples containing higher portions of seeping HC are recognized. These samples, represented by large red dots on the line graph, are closely associated with the outlined oil field at a depth of almost 3000m.

Figure 4
Figure 4 shows a large regional survey conducted in the Barinas Plains and Foothills regions of Western Venezuela with reservoir depths between 3000 - 5000m. Note and evaluate post survey drilling results shown on the map in comparison with previous GEL anomalies. The G1 well in the Guasimito structure was lost (mechanical failure); the final status of this structure is not clear at this point in time. The wells U1 and U2 are old wells; the GEL anomaly associated with these old wells prompted re-evaluation of the logs; new log data show 30m of net pay in sand reservoirs. For more information we refer to Callejon and von der Dick (2002).

Limitations of GEL and Continued R&D

A limitation exists in areas of very low HC seepage to the surface. So far, roughly 70% of existing fields and successful prospects are GEL diagnostic, which inevitably leads to some missed, good opportunities (see GEL Track Record). However, as outlined above, GEL evaluated prospects contribute substantially to increased drilling success although not all production is GEL surface gas indicated. Furthermore, the iGEL procedure has the capability to identify seepage in very low seep environments. iGEL is diagnostic for > 70% of all prospects.

Current R&D is in progress to extent the GEL application. CTE (Germany) is presently involved in a microbial surface exploration method (BEL) jointly with the University of Aachen, Germany, and a consortium of British and Greek institutions under an applied EU R&D initiative. Stacking of GEL seep data with BEL seep data will provide an estimate of the total active seepage and increase the S/N ratio further.

References:
For the complete listing of the quoted references we refer to CTI's assembled CTI Publications and SGE literature listings.
Top of the page


Web site developed and maintained by Spiderling Inc.