25TH INTERNATIONAL NORTH SEA FLOW MEASUREMENT WORKSHOP October 16-19, 2007, Oslo, Norway Flare Metering with Optics From Blue-Sky Technology to the Real World Jody Parker, Gordon Stobie, ConocoPhillips Company Ivan Melnyk, Photon Control Chip Letton, Letton-Hall Group

Abstract

During a presentation at the 2006 North Sea Flow Measurement Workshop on new and current flare metering technologies, infrared and other optical methods of gas flow measurement were referred to as “blue-sky” technology. The inference was that these technologies, whilst they might have merit, probably would not be available for some time, if ever. In reality, this is far from true, as at least one vendor of such devices has developed its technology to the point that laboratory and field testing have been carried out, and more than twenty-five units have been sold and installed into operational plants.

The purpose of this paper is to review the development, testing and deployment of the Photon Control Optical Flow Meter. In particular, the following topics will be addressed:

1. Overview of the technology, its various embodiments, its advantages and shortcomings, with a synopsis of a Canadian JIP under which development was carried out.
2. Presentation of results:
   • Flow Laboratory testing for installation effects
   • A variety of general Canadian onshore retrofit installations
   • Specific examples from the eight ConocoPhillips Canada “Real World” installations
   • An example using optical flare metering technology in Statoil
3. Detailed conclusions on the JIP, the gradual implementation of the new technology into the industry, with feedback from both users and regulators on being able to manage flare gas discharges.

1. STATE OF OPTICAL FLOW METERING: A REVIEW

Historically, optics are less well known in the realm of gas flow measurement whereas analytical instrumentation – LIDAR*, gas analyzers, etc., use the inherent features of light such as specific absorption, fluorescence or scattering which cannot be realized by any other techniques. Optical methods for measuring gas flow, or optical flow meters (OFM), use optical velocimetry, the measurement of gas flow velocity from which the volumetric flow rate can be derived. These methods can be subdivided into laser Doppler velocimeters (LDV) and optical transit time velocimeters. The latter can be divided into laser-two-focus (L2F), scintillation-based and absorption-based transit time velocimeters.

1.1 Product History

Spectron Development Laboratories conducted a study for the Gas Research Institute to determine the possibility of developing a volumetric gas flow meter based on the L2F technique in 1989.3 Although this study never resulted in a commercial device, it caught the attention of Nova Husky Research where optical methods were in use for particle sizing and monitoring the quality of filters. A project on the L2F volumetric flow meter was conducted for several years at Nova, and later at TransCanada Pipelines Ltd. (TCPL). The effort was continued in 1999 - 2000 as a joint development project with a major flow meter manufacturer.4,5 The work focused on developing a high-accuracy Optical Flow Meter (OFM) which would be suitable for gas custody transfer measurement, and in particular for the replacement of orifice meters. In 2002 Photon Control Inc., licensed this optical flow metering technology from TCPL for the purpose of its further development and commercialization in a variety of gas flow metering applications.

During 2003 Photon Controls proposed to the Canadian oil and gas industry a Joint Industry Project venture in order to mitigate the meter development costs and to gain access to ‘real world’ facilities. Unfortunately the take up on this venture was poor, and it was not until somewhat later at the 2004 Canadian School of Hydrocarbon Measurement in Calgary, Alberta, that ConocoPhillips Canada (COPC) became aware of what looked like an attractive option for flare metering. At this time COPC and Photon Control began to collaborate with the development of the meter systems for flare and vent gas applications.

COPC recognised that there was a ‘hole’ in the management of their plants and resources, and saw that local regulations would soon require them to report flare and vent quantities to a level which had been unachievable in the past. COPC were uncomfortable with their existing estimates of stack losses and wanted more accurate information in order to reduce or eliminate background gas to the flare systems. It was seen that flare/vent metering was the only way forward, and that the existing flare and vent meters at that time did not meet the Business Units’ needs as either being fit for purpose or especially cost efficient. The collaboration process continued through 2006.

1.2 Laser Doppler Velocimetry

LDV is the oldest form of optical velocimeter, and was proposed soon after appearance of the first commercial lasers. However, LDV has found little industrial or commercial application because of its high cost and the need to particle-seed the flows due the very low signal-to noise ratio. In laboratories, LDV offers impressive accuracy and the ability to measure very high velocities. This, combined with the ability to measure complex 3D gas flows makes LDV an important tool in turbomachinery and avionics applications.

1.3. Optical Transit Time Velocimeters

1.3.1. Laser-Two-Focus (L2F) Meters
Thompson first described the possible implementation of the L2F method for flow measurement in 1968.1 Schodl contributed significantly to the practical aspects of the L2F technique, but the method never went beyond flow laboratory implementation other than a few commercial L2F meters built by Polytec in the early 1980s.

1.3.1.1. Principle Of Operation
The operational principle of the optical gas flow meter based on L2F velocimetry is explained in Figure 1. Small particles which accompany natural and industrial gases pass through two laser beams focused in a pipe by illuminating optics. Laser light is scattered when a particle crosses the first beam. The detecting optics collects scattered light on a photodetector P1, which then generates a pulse signal. If the same particle crosses the second beam, the detecting optics collect scattered light on a second photodetector P2, which converts the incoming light into a second electrical pulse. By measuring the time interval between these pulses, τ, the gas velocity is calculated as

where S is the distance between the laser beams.

Figure 1. Operating principle of the L2F velocimetry

1.3.1.2. Accuracy Of The L2F Method
The linear gas velocity can be measured with high accuracy using the L2F method independent of pressure, temperature and gas composition. Using (1) above, the velocity uncertainty σv can be estimated as

where σ_d and σ_t are the standard deviations in velocity due to errors in optical spacing and lapse time, respectively. The uncertainty of the optical spacing is defined by the accuracy at which the beam spacing can be measured. For typical beam spacing d=1mm and positioning uncertainty of Δ_d=1 μm, the typical optical uncertainty would be around 0.1%. The lapse time uncertainty could be even smaller, as it is defined by the sampling frequency f_s. Smart, for example, reported a velocity uncertainty of less than 0.02% while using analog-to-digital conversion at a sample frequency of 100MHz.⁶ The number of particles effectively crossing the two laser beams, N, contributes to the velocity uncertainty as approximately . Conversion of the linear velocity measured at a single point to an average velocity, however, leads to a larger uncertainty due to the flow profile variations and turbulence. According to Schodl, the total error of the L2F velocimeter could be as low as 0.5% in a predictable profile, if the turbulence does not exceed 4%.⁷

The value N is determined by the following factors:

1. the meter itself which includes: the efficiency of the delivery and collecting optical systems, the detectability of the photodetectors , the laser power and the wavelength;
2. the purity of the gas moving in the pipe;
3. the gas velocity profile and turbulence level.

In contrast to LDV devices, L2F velocimeters usually do not require seeding because of their inherently high signal-to-noise ratio, SNR. High SNR in L2F velocimeters results from the concentration of laser light into two focal sheets. LDV signals, however, consist of multiple fringes occurring after the interference of two convergent laser beams. Photodetectors such as the avalanche photodiodes used in L2F gas velocimeters, register individual photons, which allows them to use relatively low power lasers. These mass-produced semiconductor lasers transmit from 1 to 5 milliwatts through single-mode fibers, and can be focused into narrow sheets measuring between 20 and 30 μm wide in a 2- to 6-inch pipe.

The collecting optics must be designed to collect the scattered light within as large a solid angle as possible while blocking all direct light coming from the sheet. For 2-inch and 4-inch meters, the dark-field collecting optics must block the straight light from 0 to 2.5 degrees. Light scattering efficiency is determined by the size of the particles and the laser wavelength. L2F velocimeters operated at near-IR (850nm) can measure the velocity of air with a minimum particle diameter of approximately 0.3μm.⁶ Shortening the laser wavelength reduces this minimum detectable particle size to less than 0.1 μm. During the early development of the L2F gas flow meter, particles found in a typical gas pipeline were shown to range from 1 to 10 μm.³

Lowering the gas velocity reduces the number of detectable particles. At a certain minimum velocity, V_min, the OFM cannot distinguish the organized flow from the stochastic movement of particles in the pipe due to ‘thermal stratification and other external factors’. The value of V_min establishes the minimum measured flow rate, and thus the rangebility of the meter.

1.3.1.3. Turn-Down Ratio
The turn-down ratio, or rangebility, is probably the most important parameter of any flare gas meter that is proven to be repeatable. Some manufacturers of ultrasonic flare meters claim values of V_min of 0.03 m/s and V_max of 80 m/s, yielding a turn-down ratio in excess of 2500:1. The turndown ratio of 100:1 recently claimed for a 12-inch flare meter from Instromet should perhaps be considered realistic for the class of ultrasonic flare meters. In contrast to ultrasonic meters, L2F flare meters have virtually no limits for V_max. An extreme maximum velocity, up to V_max=720 m/s, was reported during the testing of the L2F velocimeter in a recent supersonic aircraft⁸ test.

The minimum velocity for Photon Control’s OFM is defined by presence of particles - the dirtier the gas, the lower Vmin is possible. It has been shown that flow through the OFM can be measured down to V_min =0.1 m/s, as confirmed by testing in the flow loop at CEESI described below. Testing for a high Vmax is limited in larger pipes by the limitations imposed by the available flow testing facilities. L2F OFM have been tested up to V_max=100 m/s ⁹, which is used to define the L2F turn-down ratio as 1000:1.

1.3.2 Other Optical Meter Types
There are a number of other optical meter types using different principles, and these are listed briefly for completeness.

1.3.2.1 Optical Scintillation Meters
Ting-I Wang¹⁴ described an OFM whose operational principle is based on a scintillation technique, or registration of variations in refraction of the light beam caused by local fluctuations of the refractive index produced by turbulence and heat exchange in the gas. The scintillation OFM is the only gas flow meter whose performance improves with turbulence. An improved version of the scintillation OFM called the ”Laser-Two-Beam”, or L2B meter is offered by Photon Control for larger nominal bore pipes.

1.3.2.2 Optical Absorption-Based Meters
Liquid hydrocarbons and water absorb energy more than gases in the IR region. This effect was the basic operating principle of an absorption-based OFM originally proposed for non-gaseous flow applications, such as asphalt and cement production.¹²

The technique was attempted recently for flare gas measurement, although laboratory testing using simulation media of air and water droplets was able to reach only Vmin = 0.4 m/s.¹³ The principle of the IR-absorption meter is illustrated in Figure 3. Two collimated beams from IR LEDs or IR lasers cross the pipe perpendicular to the gas flow. The presence of hydrocarbons or water droplets in the flow causes changes in the optical transmission, which is detected in each channel. A cross-correlation technique is used for calculation of transit time.

An absorption-based meter requires the presence of large water droplets that discretely cross the beams amid turbulent hydrocarbons. Uniformly distributed water vapour and/or methane lead to static changes of the absorption, which can make transit time measurement very difficult. A fundamental drawback of the absorption method is that it actually does not provide averaging across the pipe. It is probably this effect which is responsible for an average of 15% error being recorded during the lab testing of the IR absorption meter reported by NEL13 at the NSFMW in 2006.

1.3.2.3 OFM Based On Sagnac Effect
Similarly to ultrasound, light passing along and against the gas flow will have different phase velocities which are related to the gas flow velocity. This difference can be detected using the Sagnac effect; the method was first described and demonstrated by Blake.¹⁴

Although encouraging laboratory data has been demonstrated15, the Sagnac OFM has never been fully investigated. The meter will be sensitive to vibration of the pipe and flow turbulence; also, one can expect significant beam deviation at a long optical distances due to the temperature gradient in the pipe.