Tracers which pass all the batch experiments with acceptable marks advance to the dynamic tests where their flooding properties are examined in core flooding experiments. This is the last laboratory test before final qualification in field pilot tests.

Injection of a radioactive tracer into a hydrocarbon reservoir at high pressure entails handling of materials with properties that are unfamiliar to most laypersons. As radioactive material in quantities used for injections is potentially dangerous, it is important that it be treated properly. In an injection project, safety considerations should have already been taken into account when the tracer is selected and the quantities fixed. During transport and storage at the well head site, the safety aspect should be borne in mind and safety precautions should be taken according to international rules and regulations.

Although some limitations of the original models persist in PORO (flow in parallel no communicating areally homogeneous layers, constant water saturation, Sw conservative tracer travelling jointly with the bulk water flow and steady state flow) it constitutes an advancement in the sense that it incorporates very common aspects of the reservoirs, such as non-regular unbalanced patterns, anisotropy and sealing faults. Also, it is important to highlight that the model may be easily extended for incorporating adsorption, radioactive decay and tracer partitioning in the same way as in the Abbaszadeh-Dehghani and Brigham model

[35, 36] and also it can incorporate the transport of tracers in stationary gas flow [40].

2.3.1. Fluid sampling in production wells

Sampling must be carried out according to the planned schedule and in most cases should start immediately after injection.

Sampling frequency will depend on the basic understanding of reservoir production dynamics. The decision to stop sampling is made jointly by the end — user and tracer team based upon the development of the tracer production curve and cannot be planned a priori. Generally, sample volumes of 0.5—1.0 L are preferred. This enables repeated analysis of the same sample if desirable. This possibility may be important for quality assurance of the analysis. In the case of water cut, preferably less than ~10% of the water sample is collected at the separator equipment point where the majority of the oil is removed.

In the case of too much oil, the water has to be recovered from a complex oil-water emulsion, and the quantity of water recovered may still be too little to obtain a high quality analysis with the best detection limit. For sampling in geothermal wells this problem does not exist.

A typical sampling schedule is introduced below. Tools for well head sampling include:

• Personal protective equipment (safety glasses, gloves, personal H2S

monitor and respirator (if required by local rules), etc.);

• Crescent wrenches;

• 20 L bucket for oil and water spills;

• 1 roll of electrical tape for sealing up the bottle cap;

• 1 L plastic wide mouth sample bottles;

• Absorbent pads (for oil spills);

• Rags (for water spill);

• Felt tipped permanent marker for writing on bottles;

• Sampling record book.

The sample is marked and identified by the well number, date of collection and, preferably, by the initials of the sampling engineer. The actual schedule of sampling and the method used is field dependent. Different methods may be considered:

• Manual sampling carried out by the operator;

• Automatic sampling at the well head;

• Continuous sampling;

• Downhole sampling.

1.3.1. Introduction

Radiotracer applications for interwell communication studies were carried out in two oilfields in Pakistan. The study area consists of two oilfields (oilfield 1 and oilfield 2) and is shown in Fig. 58.

1.3.2. Tracer test in oilfield 1

The study is being carried out in the Fimkassar oilfield, which is operated by the Oil and Gas Development Co. Ltd. This oilfield is located in the Potowar Basin, about 100 km southwest of Islamabad. The field has two production wells (wells 1 and 2) and the third well (well 3) is used as an injection well for water flooding. A fourth well was drilled but it was ‘dry’ and was capped. The pattern and the location of the wells are shown in Fig. 59.

Objectives

The main objectives of the study were to:

• Determine breakthrough time between injection and production wells;

• Assess the contribution of injected water towards the production well and investigate the presence of quick channelling between the injection and production wells;

• Determine the swept volume of the reservoir by injected water;

• Assess the efficiency of injected fluid to increase the reservoir pressure;

• Assess the relative contributions of injected water and formation water in the produced water.

Monthly sampling of producer wells and the injection well was carried out and samples were analysed for tritium (3H) and stable isotopes (2H, 18O) using liquid scintillation counting and isotope ratio mass spectrometry, respectively. Radiotracer and stable isotope data were processed and analysed and these data are displayed in Figs 60 and 61.

The work related to this oilfield is complete and the results are as follows:

• Breakthrough time is 252 d.

• Water produced in well 1 has an 85% contribution of fresh injected water.

Inter-Well Tracer Test (Fimkassar Oilfield, Pakistan)

Time after Radiotracer Injection {Days)

FIG. 60. Tracer response curve and tracer recovery (Fimkassar oilfield).

• Mean residence time of tracer (water) is 1367 d.

• Mean injected water volume is 2 497 864 m3.

• Mean produced water volume is 460 711 m3.

• Maximum and mean velocities of injected water are 10.3 m/d and 1.9 m/d, respectively.

• Tracer recovery is ~69%.

Considering a mean produced water volume of 460 711 m3 and 69% recovery of radiotracer from well 1, the mean swept volume is determined as 317 891 m3. This is the average volume of the reservoir swept by injected water.

• After initial reduction in breakthrough time, the injection water has swept a large volume and there appears to be no channels connecting the injection well and production well 1.

• No injected tracer was detected in well 2.

• The tracer response shows that the water flood regime was not managed properly in the initial stages that changed the hydrodynamics of the reservoir and affected the production.

• The water in the formations is meteoric water, recharged from northern parts of the country.

III.1.5.1. General

Since HTO is the most frequently used water tracer and, in addition, applied in the highest quantities per injection (in becquerels), it is used here as an example on discussion of the environmental impact of such operations.

Sea water contains low concentrations of practically all radioactive nuclides present globally. The radionuclide [1]H is also present in sea water. In the North Sea, the concentration of 3H is in the order of 1 Bq/L. During injection, there will normally be no release of tracer into the sea from offshore installations. However, in case of an abnormal situation arising involving spillage of tracer, then the spilt tracer will be dispersed into the sea. The impact on the environment caused by this type of tracer discharge therefore has to be evaluated.

111.1.5.2. Impact of an accident during injection

(a) Worst case from radioactive tracer discharge into the sea

The worst case impact on the environment will occur if a whole portion of tracer has to be discharged into the sea.

(b) Impact on the European population from HTO discharge

The following is an example on how this has been evaluated for a typical North Sea situation.

The report NRPB-R109 from the British National Radiological Protection Board, A Model to Calculate Exposure from Radioactive Discharges into the Coastal Waters of Northern Europe, contains a suitable model for calculations of dose commitments to the people in the region.

111.1.5.3. Dose commitment from 3700 GBq of HTO

From the NRPB-R109 report, it is possible to calculate that a discharge of 1 GBq in one year gives a collective intake of 57.07 Bq over a period of 50 years. From this, 2000 GBq of HTO gives a collective intake of 211 159 Bq. A discharge of 50 MBq HTO gives a dose commitment equal to 1 mSv. Thus, the collective intake of 211 160 Bq originating from a discharge of 2000 GBq gives a dose commitment of 4.2 x 10-6 mSv.

The residence time distribution method is used largely for modelling classical chemical engineering vessels and reactors. The classical chemical engineering approach may be adapted for oil field tracer experiment interpretations. Some attempts have been made using the residence time distribution software (DTSpro) to verify how it can be used for non-boundary systems and low tracer recovery. In practice, the software package may be used but it is not well adapted to the problem. The elementary bricks available are representative of simple flow: dispersive flow and dispersive flow exchanging with the porous zone of lower velocity, but the parameters are not directly correlated with the usual parameters used by oilfield engineers. Moreover, the artifact used to take into account the loss of tracer is complex; it is not recommended for such an application. On the other hand, the recent approach of the compartmental model derived from both CFD and residence time distribution is promising. It consists of structural and functional descriptions of the studied structure (CFD is only a structural description with a mesh and DTSpro is used as a functional description of the structure with a network of elementary behaviours).

Figure 122 shows an academic example of such a model. The advantage of such a model is that it is derived from results from several models (PORO for the streamline network, CFD, information about fracture). In a first step, the user should define an elementary ‘slice’ based on the internal structure of the oilfield, the exchange flow of major importance (radial dispersion, convection) and the boundary layer continuity rules. It is then necessary to calculate the flows
between the different parts and the number of slices simultaneously by an iterative process and by fitting the tracer response.

2.3.4.1. Correcting raw data

Raw data from the analysis should be corrected before evaluation, interpretation and reporting to the end user. Examples of corrections are:

• Correction for background radiation;

• Correction for radioactive decay;

• Corrections for chemical yields and counting efficiency;

• Conversion to concentration units required by the end-user (tritium units (for tritium labelled tracers only), Bq/L, Bq/g, etc.).

2.3.4.2. Data availability

Analytical results should preferably be compiled in a database system that is accessible by the end user at any time.

2.3.4.3. Recording data

Data should be compiled continuously along the timeline as soon as a new analysis has been performed and presented in tables with a clear identification of

the reservoir and the well label in the table heading. Data may, in addition, be presented in the graphs or curves (production profiles for instance) for easier evaluation and interpretation (improved perception).

2.3.4.4. Reporting

Various types of report may be required:

• Report of the experimental activity describing the whole process up to, and including, tracer injection;

• Brief reporting (e. g. by email) upon completion of the analysis of each received batch of samples;

• Final report upon completion of the whole project which presents the full range of data and summarizes the major findings which can be extracted by a simple qualitative evaluation of the data before any quantitative interpretation.

The experimental laboratory intercomparison tests (‘round robin’ test) reported here were twofold: (i) analysis of HTO in different waters and (ii) analysis of HTO plus 14CH3OH in mixtures in different waters. The first test sequence was organized by Norway while the second was arranged by Vietnam.

11.1. LABORATORY INTERCOMPARISON TEST ON HTO ANALYSIS

11.1.1. Preparation of samples

The following series of samples were prepared at IFE, Norway:

• Series 1, samples 1A-1E: Distilled tap water with added and calibrated quantities of HTO. Sample volume was 100 mL each.

• Series 2, samples 2A-2E: Distilled tap water with a quenching agent added and calibrated quantities of HTO. Sample volume was 100 mL each.

• Series 3, samples 3A-3E: Artificial Gullfaks (Norwegian oilfield) formation water with added and calibrated quantities of HTO. Sample volume was 250 mL each.

• Series 4, samples 4A-4E: Real formation water (Tordis oilfield, Norway) which has been in contact with ‘live’ oil. A calibrated quantity of HTO was added. Sample volume was 250 mL each. For the concentration of some main ions, see Table 17.

The samples were calibrated with certified Wallac and Packard secondary standard tablets using Quantulus 1220 liquid scintillation counting equipment. Six packages of 20 water samples each were shipped to the various participating laboratories on 18 May 2006. Results were requested by 15 September 2006.

111.5.1. Tritium and 14C analysis in alcohol

The presence of tritium and 14C in alcohol in a water sample taken from the oilfield cannot be analysed directly by liquid scintillation counting; it needs to be purified using the distillation process described below.

111.5.2. Equipment and reagents

• Round flask 500 mL, connection size 24/40;

• Fractional distillation Vigreux column 31 cm long, connection size 24/40;

• Dean & Stark collector sized 10 mL;

• Cooling system;

• Heater;

• Magnetic stirrer;

• Liquid scintillation counter with standard 10-20 mL vial;

• Methanol and toluene reagent grade;

• Cocktail Instagel or Ultima Gold.

111.5.3. Procedures

(1) Sample treatment: Add 10% v/v of toluene and extract water by funnel.

(2) Place water sample in standard vial with appropriate cocktail to count 3H and 14C directly on liquid scintillation counter at dual label mode. Calculate tritium (HTO) activity with correction for 14C in methanol contribution.

(3) Add 3 mL of methanol and 1.3 mL of toluene to water sample in the round flask.

(4) Add a magnetic stirring bar to the round flask on the distillation system.

(5) Heat the sample at low power to distil for 5 h.

(6) Collect the distillate (about 4 mL) into the counting vial.

(7) Add cocktail and count by dual label mode with liquid scintillation counter.

(8) Calculate the 14C activity (from 14C-MeOH) using the appropriate correction for HTO influence.