Sampling systems – the cash registers of the gas supply chain
June 23, 2014
First published in Platform Oil, Gas and Renewable Technology Review, issue 171-2014
Wherever a business sits in the supply chain, certainty over the value of the gas passing through its process is crucial. Roger Brown of Thyson Technology explains why effective integration of sampling technology is vital in achieving accuracy and efficiency.
On its journey from reservoirs deep under land or sea to the point where it is consumed by the end user, natural gas can change hands dozens of times.
Every party along the supply chain pays for the gas it receives and gets paid for the gas it passes on, so the ability to measure the value of the material accurately – a process known as fiscal metering – is critical. Where high volumes of gas are involved, any inaccuracies can add up to financial errors worth thousands of dollars per day.
This is especially crucial for transportation businesses, which can operate on razor-thin margins of a matter of pennies for every million cubic feet of gas that passes through.
Two values are needed to determine the value of gas – mass and calorific value (CV). The second of these is a measure of the heating power of the gas and something that varies depending on its composition.
To establish the mass of gas passing through the transfer point, a mass flow meter is placed in the pipeline and data from this is constantly logged to give a precise calculation of the mass that passes through in a specific period of time. There are a number of methods in common use for measuring the mass flow of gas including differential pressure measurement, turbines and ultrasound-based systems.
However, to determine the value of the gas, the mass data needs to be combined with analytical data on its composition, and this can be a challenge as the composition of natural gas can fluctuate.
CV analysis relies on small samples being extracted from the flow to give a snapshot every two or three minutes of the composition of a constantly flowing gas at a particular point in time.
Gas chromatography is the most commonly used approach, a method that determines the calorific value by separating out the constituent compounds in the gas to measure its composition.
Thanks to advancements in micro-processing, chromatograph technology has developed rapidly in recent years, and the usability of analyser units, the range of readings they are able to take and the frequency with which these can be delivered has improved significantly while their physical dimensions have decreased.But these devices are not plug-and-play, and they require a sampling system that will work with the specific pipeline in question. However high-performance an analyser might be, the results will only be as good as the sampling system within which it is operating.
The system needs to deliver samples of the correct size, at the correct temperature and pressure and with the correct transfer time. This essential in providing data with high enough resolution to accurately measure the changing composition of the gas.
Ensuring representative sampling
If the composition of the gas reaching the analyser differs from that in the process, this will result in an inaccurate reading which could have significant consequences in terms of over or undervaluing the gas.
A common cause of non-representative samples is the need to reduce the pressure of the gas significantly from that in the pipeline to that required by the analyser, which is normally atmospheric pressure.
This drop in pressure can cause heavier hydrocarbons with higher dew points to condense out of the gas – a phenomenon known as ‘heavy-end drop out’. If this is allowed to occur, the CV measured by the analyser will be significantly lower than the real value of the gas in the process and the gas will be under-valued as a result.
The samples generated by the pressure let-down system therefore must be tested periodically by comparing them with manual spot samples taken directly from the pipeline. When carrying out this operation, it’s important to ensure that the sample from the pressure let down system and the spot sample were both extracted from the process at the same time, so any lag time in the sampling system must be known and taken into account.
If heavy-end drop out is detected, more heat must be applied during the pressure let-down process to ensure heavier compounds remain in the gas.
As well as ensuring the sample is representative of the gas being tested, the analyser itself must also be checked for accuracy by feeding in a gas sample with a known, lab-tested composition, checking that the results generated agree with expectations and making calibration adjustments if necessary.
In some circumstances, these challenges are compounded by the need for additional analyser systems working alongside the CV chromatograph for components in the gas which it is unable to detect, such as moisture or sulphur. These requirements further increase the complexity of the sampling system.
The high masses that pass through processes across the gas supply chain means tiny inaccuracies in sampling can add up to significant financial losses.
To illustrate this point, let’s imagine a process is handling 2.8 million cubic metres of gas per day and the CV measurements provided by the GC system differ from the actual value by just 0.35 per cent – not an unrealistic figure.
This modest error would result in a daily measurement discrepancy of 486,766 megajoules (MJ) or 135,219 kilowatt hours (kW•h) which, based on average recent gas prices, means the gas would be under-valued by £3,325 per day and £99,751 per month.
Fundamentally, a sampling system consists of an arrangement of pipework, valves and small flow meters that transfers the sample, ensures it is in the right condition when it arrives at the analyser and cleans out the system once the operation is complete ready for it to be repeated. But the simplicity of this statement belies the complexity of the systems and the number of individual operations they carry out each time it delivers the sample.
If an operator was required to run the process manually each time a sample needed to be taken, it would be a full-time job. Instead, sample systems are designed to run the process end-to-end, automatically, and this can involve complex automation systems.
More advanced systems will run completely autonomously, further freeing up operators’ time, but this requires the system to operate hand-in-glove with a control system that will run it and take the measurements automatically. Of course, reliability of this system is critical, as regular breakdowns requiring interruptive maintenance or causing lost data would more than defeat the object in terms of cost to the operation.Made to measure
From project to project, the requirements for sampling systems vary significantly, whether because of the make-up of the gas being measured, differing requirements for accuracy and frequency of measurements and more pragmatic concerns like the geometry of existing or planned pipework and any space and weight restrictions.
For this reason, there is no one-size-fits-all solution and, while turnkey solutions can be applied in some circumstances, the installation must always be based on an intimate knowledge of the specific requirements of the scheme in question.
Ultimately, confidence in the accuracy of calorific values is something that the gas industry cannot live without, but the changing nature of the upstream supply chain has created new challenges for system designer in terms of automation, budgets and space considerations.
As extraction projects get more light-weight and budgets continue to get squeezed, the onus is on sampling system engineers to keep pace with developments in analyser technology to deliver systems that are affordable, compact, automated and highly reliable without sacrificing accuracy or safety.
Roger Brown is engineering director at Thyson Technology