GB2580622A

GB2580622A - Method to improve the pressure variation tolerance of gas measurement devices

Info

Publication number: GB2580622A
Application number: GB1900645.1A
Authority: GB
Inventors: Dinsdale Steven
Original assignee: Cambustion Ltd
Current assignee: Cambustion Ltd
Priority date: 2019-01-17
Filing date: 2019-01-17
Publication date: 2020-07-29
Anticipated expiration: 2039-01-17
Also published as: GB2580622B; GB201900645D0

Abstract

A sample probe 1 takes sample from a source of emissions such as an engine, which may be at sub-ambient or super-ambient pressure. The sample probe supplies sample gas to a gas analyser 6 via a capillary tube 5. Downstream of the entrance to capillary tube is a short continuation of the sample probe. Downstream of the continuation is a diffuser 7 which recovers static pressure lost due to wall friction in the continuation. Downstream of the diffuser is a chamber 3 held at constant pressure connected to a vacuum source 4. The combination of the constant pressure chamber, the short distance from the capillary to the chamber, and the diffuser improve pressure isolation of the end 8 of the capillary tube from sample pressure fluctuations, and thus also isolates the flow in the capillary tube to the analyser from sample pressure fluctuations. This ameliorates inaccuracies in the analyser flow rate changes due to pressure fluctuations at the sample source. The analyser may be flame ionisation detector (FID) or a chemiluminescence detector.

Description

Method to improve the pressure variation tolerance of gas measurement devices This invention relates to fast response gas analysers, where isolation of the component that converts the gas composition to a measurable quantity (e.g an electric current, or light signal), from the source pressure is critical to the accuracy of the measurement.

In a prior invention [US5073753] a device was described which addressed this requirement as applied to a Flame Ionisation Detector (FID) for measuring the concentration of hydrocarbons (HC). Figure 4 from that patent is reproduced schematically as Figure 1. A sample probe (1) samples exhaust gas from an engine, either directly from the cylinder, or from the exhaust manifold. A chamber held at constant pressure (3), connected to a vacuum source (4) is connected to the downstream end of the sample probe (1). Immediately prior to the constant pressure chamber (3), a thin capillary tube (5) draws off gas from the sample probe, to supply a HD analyser (6). In that patent, attention is drawn to the need to have tube element (2) as short as possible so that the pressure drop along it (due to wall shear) is small and results in an acceptably small influence on the flow rate in the capillary tube (5) delivering sample to the FID itself This is important as the FID is sensitive to changes in sample flow delivered to it, and once calibrated at a certain sample flow, requires that flow to remain as constant as possible to maintain accurate results. Elements (1), (5) and (2) form a similar shape to a letter "T" -so this arrangement is sometimes described as a "T".

Modern turbocharged engines are subject to a wider range of cylinder and exhaust pressures than the engines of 20 years ago. The present invention improves on the prior invention by adding a diffuser or diffusers to the pressure isolation system, in order to maintain pressure isolation when performing measurements in modern engines. A subsonic diffuser is an element in which the cross-sectional area gradually increases along the direction of flow. This increase in cross sectional area causes a fall in the velocity of the gas, and thus by the Venturi effect, a rise in pressure. This rise in pressure can be harnessed to counteract the fall in pressure in tube element (2) (a continuation of the sample probe beyond the extraction capillary) caused by wall shear. The diffuser is designed such that its pressure recovery characteristic balances closely the undesirable pressure drop generated by wall shear. Hence the addition of a diffuser or diffusers allows better isolation of a gas measurement from the gas source pressure variation.

Examples of the invention will now be described by referring to the accompanying drawings.

Figure 2 shows in cross-section an example of how the invention may be implemented in practice. In this example, the original invention shown in Figure 1 has been modified to include a diffuser (7). By an appropriate choice of diffuser length and included angle (though the geometry of the diffUser might be of more complex geometry) the pressure changes at point (8), resulting from sample pressure changes at point (9), are significantly reduced. In addition, the invention allows the length of tube (2) to be chosen more flexibly. For example without the present invention, this length must be as short as possible, but this might lead, at some conditions, to regions of reverse flow in tube (2).

The prior invention US5073753 describes the use of a diffuser as an alternative arrangement to the "T" arrangement described above, for example in its figure 1 B, and column 4, lines 5765. However, in that embodiment of that invention, the diffuser is placed upstream (rather than downstream) of the sample capillary leading to the FID and has a completely different function that diffuser's role is to reduce the velocity of the whole sample flow to such an extent that its dynamic pressure head becomes negligible. There is no "T" arrangement of elements (I), (5) and (2) in that embodiment of the prior invention -indeed element (2) is not present. In the current invention, the diffuser is placed downstream of the sample capillary, with the objective of correcting for a less-than-perfect aspect of the "T" arrangement. As described in US5073753, the "T" method aims to achieve isolation from sample pressure variation by the "pitot-static" method, in which a portion of the sample is taken at right-angles from the main flow, so that the (varying) dynamic pressure of the sample flow has negligible effect of the near-constant flow passing to the HD. The current invention describes a method of correcting for the less-than-perfect characteristic of the "T" arrangement, which is that element (2) creates a certain amount of skin friction, which means that the pressure at point (8) is slightly higher than that in the constant pressure chamber (3). The addition of a suitable diffuser (7) corrects for this phenomenon. Thus, the diffuser of the current invention does not seek to render the flow in it so slow that its dynamic head is negligible (as in US5073753), but to make a slight correction to the behaviour of a device in which a diffuser is not the primary modus operandi. In the prior invention, In Figure I B of US5073753, a diffuser is the modus operandi.

There may be a plurality of gas analysers connected to a single sampling system. For example, Figure 3 shows an embodiment where two capillary tubes (5) are connected to the sample probe (1) prior to the diffuser (7), each supplying pressure isolated sample flow to separate gas analyser (6). The gas analysers (6) are not limited to being FIDs for hydrocarbon detection. Any gas analyser which is sensitive to the rate of sample delivery would benefit from this pressure isolation method, for example Chemiluminescence Detectors (CLDs) for nitrogen oxide, or Laser Induced Fluoresce (LIF) detectors for nitrogen dioxide.

Figure 4 shows, in cross section, an alternative embodiment where a further gas sensor device is added downstream, i.e. in series; in this example an optical absorption type. The constant pressure chamber (3) is illuminated by a source of infrared light (10). Before passing through the chamber (3), the light source is attenuated by a filter or filters (11) which closely match the infrared absorption fingerprint of the gas or gases of interest. After passing through the chamber (3), via windows (12), the light enters an infrared detector or detectors (13). Changes in the concentration of the gas or gases of interest cause corresponding changes in the level of infrared light absorbed and thus a change in the level of electrical signal from the detector or detectors (13) which is related to the gas concentration.

For example, such an optical absorption gas sensor could be used to measure the concentration of carbon monoxide, carbon dioxide, and/or water vapour. With the use of multiple filters and multiple detectors, this optical absorption gas sensor can measure a plurality of different gas species simultaneously. Given as stated above it is possible to have multiple flow sensitive sensors (6) prior to the absorption sensor in the constant pressure chamber (3), it is possible to simultaneously measure most of the gasses emitted by internal combustion engines using a single sample line (1). In this arrangement (Fig 4), if the optical chamber itself is held at a constant pressure then a further diffuser is not required. however in some embodiments it may be useful to have a second constant pressure chamber downstream of the optical constant pressure chamber (3), in which case the arrangement shown in Fig 5 may be used to advantage.

In this case an additional diffuser (14) isolates the optical sensor from pressure losses due to the wall shear in the piping leading to the second pressure controlled chamber (15). In addition, it is clear that the second diffuser also acts to create a near-constant pressure at the exit of the first diffuser, which is required for the proper operation of fin this embodiment) the FID sensor.

Figure 6 shows comparative data from the sampling arrangements embodied in Figures 1 and 2. A test gas of 5% propane by volume is sampled by the sample probe (1), and the pressure of that gas at point (9) varied over a range of 500 to 3500 mbar absolute, that is to say, approximately -500 mbar sub-ambient, to +2500 mbar super-ambient. The pressure deviation at point (8) was measured, as was the deviation of the hydrocarbon concentration from 5% propane. The plot shows that without a diffuser (as in Figure 1) deviations in pressure and errors in hydrocarbon concentration are both strongly correlated with sample pressure, whereas with a diffuser (Figure 2), the deviation or error is close to zero across a sample pressure range of 500 to 3000 mbar absolute.