System providing close analysis of deep-lying fluids
Philippe UngererThis apparatus has been used to investigate pure compounds, binary mixtures and real fluids from the North Sea
Institut Francais du Petrole
In the North Sea, production is scheduled from deep reservoirs at temperatures up to 190°C and pressures up to 110 MPa. This brings with it important challenges for predicting the properties of reservoir fluids, both from an experimental and a theoretical standpoint.
In order to perform fluid studies for these reservoir conditions, IFP has developed a mercury-free, high pressure apparatus that can be used to observe phase transitions, a phase sampling device, and viscosity determination by the capillary tube method.
Hercules, as the apparatus is designated, is designed for characterizing deep-lying crude oils or natural gases. Its maximum working pressure is 150MPa, and the allowed temperature range is 255-470K. The system comprises two piston cells equipped with sapphire windows so that phase changes can be visualized clearly while cell volumes are known precisely.
The two piston cells, which are identical, are located inside a large air bath and are activated by motor-driven screws, the motors being outside the heated zone. The air bath can accommodate two positions of the cells, either opposing one another or parallel (with both heads either up or down).
Maximum volume of each cell, with the piston in its extreme position, is close to 85 cc. Dead volumes are minimized by the design both of the pistons and the valves. Nevertheless, the minimum volume is around 3 cc.
Two sapphire windows allow inspection of the interior of each cell, which is illuminated by an optical fiber system. An endoscope and a small camera, located outside the heated zone, are used to display the cell's interior on a TV screen, with very good resolution (details as small as 0.1 mm can be visualized). Consequently, the vapor-liquid interface can be set with very high precision, allowing phase volumes to be measured to within +/- 0.02 cc.
Pressure measurements are performed using one pressure transducer (Dynesco 200MPa) on each cell, the membrane of which is integrated into each cell's inner wall. The sensitive element is located outside the heated air bath, as pressure is transmitted from the cell membrane using a sealed, low compressibility hydraulic fluid. In addition, a third pressure reading is made using another transducer (HBM P3MB) at the end of a tube (also outside the heated air bath) which is regularly calibrated using a dead weight gauge. The accuracy is +/- 0.02MPa.
This transducer is also used as a reference for non-linear calibration of the other two transducers in the cells. It is normally disconnected when performing measurements, so that no segregation (such as condensation or paraffin crystallization) can occur due to ambient temperature in this part of the circuit. When working at a fixed temperature, an accuracy of +/-0.05MPa is achieved over the whole pressure range.
Temperatures are measured by two platinum resistors in each cell which penetrate half of the cell wall near the upper and lower ends of the experimental volume. They have been calibrated with a platinum probe connected to a high precision thermometer. Reading accuracy is +/- 0.1K.
Heating and cooling of the cells and connection tubes is performed by air venting. Temperature control is achieved with standard electronics, with a reference platinum resistor on the air bath wall. This control is stable to +/- 0.1K, but there are also temperature heterogeneities within the air bath - recorded temperatures are slightly greater in the upper cell than in the lower. These heterogeneities increase as the temperature rises, but remain below +/- 0.5K at 374K.
Piston motion is controlled by a PC. The electric motor's high resolution control allows highly reproducible volume readings with liquids - better than 0.01%. However, corrections must be made for the influence of valve closing, tube volumes, and cell expansion. Isothermal cell expansion is particularly significant at high pressure and requires careful calibration with water.
Sampling device
In order to sample a phase in a two-phase system without segregation, it is essential that pressure be kept constant. As no reliable method is available for sampling very small volumes (around 1 cu mm), IFP opted for a larger volume with a small detachable variable volume cell (maximum 2 cc). The piston of one of the main cells is moved at the same volume as the piston of the sampling cell is withdrawn, so that pressure remains constant during sampling.The valve on the sampling cell is then closed so that the filled cell can be detached through an opening in the wall of the air cabinet. As that sampling cell is light (about 0.5 kg), it can be weighed with reasonable accuracy to determine fluid density under sampling conditions. Its contents may then be flashed and analyzed.
Determining the viscosity of gases is difficult with the standard method used for liquids, such as the rolling ball technique or the falling sinker method. The low viscosity of HP/HT gas condensates (an order of magnitude less than crude oils) allows the object to fall rapidly, making viscosity measurement difficult with these techniques.
With Hercules, viscosity is determined by measuring the pressure drop when forcing the fluid through a capillary tube with a known flow rate. Filters are placed on both ends of the tube to prevent clogging. Absolute pressure sensors are employed, although the signals they produce are subtracted before being amplified by an HBM electronics system, in order to improve the signal/noise ratio.
This apparatus has been used to investigate pure compounds, binary mixtures and real fluids from the North Sea, up to 190°C and 120 MPa. At high pressures, very high compressibility coefficient values are found for HP/HT gas condensates, exceeding 2.0. Practically, this means that at these pressures, fluid compressibility is more like that of a liquid than a perfect gas. This has important consequences for production because recovery during primary depletion is less than would be expected if classic behavior is applied.
Reference
"Measurement and Prediction of Volumetric and Transport Properties of Reservoir Fluids at High Pressure," Progress in HP/HT Fields Conference, IBC UK, Aberdeen, May 1998.Copyright 1998 Oil & Gas Journal. All Rights Reserved.
