Calorific value: fiscal measurement in the multi-gas era

Authors: Konrad Domanski, Product Manager Gas Metering Solutions, Sensirion and Andreas Rueegg, Senior R&D Engineer, Sensirion

Natural gas consumers have traditionally been billed for their energy usage measured in kWh. However, the gas meters used to measure this consumption are only capable of measuring the volume of gas passing through them. The vast majority of the meters in use today are not capable of measuring volume while taking into account the required corrections for variations in temperature and pressure (these are applied on a statistical basis by the utility companies). To arrive at energy consumption, gas distributors need to know the standard volume as well as the calorific value (CV) of gas that each consumer receives:

Energy = Standard volume * Calorific Value

Today, this problem is solved by installing gas chromatographs (GCs) at injection points in the grid and making general assumptions about the temperature and pressure of the gas delivered to consumers. Likewise, it is also assumed that all consumers beyond the injection points receive gas of identical quality. Traditionally, the CV would vary from 30 MJ/m3 for L gases to 45 MJ/m3 for H and LNG gases. In practice, this range is generally smaller, because only one type of gas is distributed in a given territory. For most H gases, the CV falls in the range of 37 to 43 MJ/m3

By separately measuring volume and CV, the gas meter can be simplified. This has historically been the most cost-effective solution. OIML R140 prescribes three accuracy classes for CV measurement: A (0.5 %), B (1 %) and C (2 %). Currently, only instruments of class A or B may be used for billing purposes in most countries.

The current trend in the global energy supply is to replace fossil fuels with renewable energy. In the context of natural gas, this transition is expected to take the form of gradually replacing natural gas with biomethane and hydrogen, before potentially switching to pure hydrogen. Consequently, many countries are now publishing their hydrogen strategies and adapting policies to prepare for this transition; hence the requirement for all boilers sold in the UK from 2025 to be hydrogen-ready. At the same time, multiple pilot projects are being launched around the world to test the readiness of networks for the arrival of renewable gases – especially hydrogen, whose properties significantly differ from those of natural gas.

The production of renewable gases is set to be much more decentralized than the current gas supply. Biomethane can be produced from biomass at wastewater treatment plants, landfills or large farms. Hydrogen can be produced at renewable power plants as an energy storage medium, from biomass or even by small electricity producers, who convert excess energy into hydrogen by means of water electrolysis. The introduction of these renewable gases into the grid is predicted to greatly multiply the number of injection points and lead to diverse CVs of the gas distributed to the end-users.

A 2020 study conducted by the Association of Italian Meter Manufacturers (ACISM) investigated the calorific value of gas close to two biomethane injection points in Italy. Due to its increased CO2 content and lack of complex hydrocarbon gases, biomethane typically has a lower CV than natural gas. The authors found that the CV measured in the vicinity of the injection points was on average 5 % to 6 % lower than the reference value used for billing customers at those locations. These results indicate that the current approach of very precise class A or B (0.5 % or 1 %) measurements at city gates, which are distant from the end-users, no longer results in fair billing of users located close to injection points of renewable gases. As the number of these points increases, and especially as hydrogen is blended in (the CV of hydrogen is only one third that of natural gas), the problem becomes ever more severe and impacts more users.

One solution to this problem is to introduce a GC at all new injection points and to model the CV of the gas distributed across the network. Depending on the number of new injection points in different territories, this approach may be sufficient. However, at the cost of approx. €25,000 per unit and additional operating and maintenance costs, the scalability of this approach is limited.

Another possible approach is to punctually measure the CV of gas in the grid and to combine it with data streamed from existing smart meters (flow, temperature, pressure and possibly other gas properties) to simulate the CV at each consumer. Such an approach is still awaiting a successful demonstration and would require MID certification to be allowed for customer billing.

The most comprehensive solution would be to measure CV significantly closer to the user and past any injection points. Ultimately, the meters themselves can be tasked to measure energy consumption directly. A combination of several approaches is also imaginable: for example, combining punctual CV measurements from GCs with readings from less accurate but significantly less expensive energy meters in order to narrow down the potential gas compositions for and refine the accuracy of the latter. A careful cost-benefit analysis of the different options will likely determine which solution will be adopted.

With over 6 million meters deployed by 2021, thermal-mass technology is on the rise for gas metering applications. The basic principle of thermal-mass flow measurement involves measuring heat distribution around a small heater while gas is flowing. Thermal-mass meters output the gas volume flow, corrected for temperature and pressure variations, and other gas parameters that are useful for diagnostics of the meters and the network (gas properties such as thermal conductivity or diffusivity, absolute pressure, etc.). The technology has been successfully demonstrated to work with natural gas, hydrogen blends, pure hydrogen and biomethane.

The measured thermal properties can be correlated to the CV of the gas and to the hydrogen content in the blend. Sensirion’s thermal gas meters can already estimate the CV of gases distributed in a network with 2 % accuracy. Figure 4 shows the results of a) measurements and b) a simulation of CV measurement accuracy across a wide spectrum of natural gas compositions (without added hydrogen). The dashed lines represent the accuracy limits of OIML class C. For comparison, Figure 4 also shows c) the simulation with an ultrasonic sensor including ideal temperature and pressure compensation. Over 1,000 real gas mixtures were used for the simulations. These included H, L and E natural gases, LNG mixtures and biogases. Biogases here are defined as a binary or ternary mixture of CH4, CO2 and C3H8, where the latter two components together constitute up to 30 % of the gas. The measurements in Figure 4 a were taken with various H and L gases (TGH or TGL in the legend).

Arguably, the ability to measure the CV of the gas at the meter level is most desirable once considerable amounts of hydrogen and biomethane are injected into the grid. Sensirion’s thermal mass gas meters are already able to accurately estimate the CV for LNG gas blends including up to 30 % hydrogen (1 %, OIML R140 class B; Figure 5). However, for gases of other families blended with hydrogen, the addition of another sensor is needed to meet similar performance.

By combining thermal characteristics measured with today’s off-the-shelf thermal-mass meters and an additional gas property (correlation based on three gas properties), a precise correlation can be achieved for a wide range of gases blended with up to 30 % hydrogen. The same can be achieved for pure and near-pure hydrogen (95 % hydrogen mixed with up to 5 % of N2 or CO2 impurities).

One such additional sensor can be ultrasonic, providing measurement at the speed of sound. It is the most intuitive solution, as the necessary technological building blocks are already available off-the-shelf. While it is possible to meet the requirements of OIML R140 class C with this combination, it requires that the compatible gas compositions be restricted. It remains to be seen whether such restriction is acceptable in practical applications. 

An ideal set of gas quantities used for correlating to CVs should be as closely related to the CV of various gases as possible while exhibiting the lowest possible correlation to each other. Sensirion investigated a set of different parameters for the task of complementing thermal conductivity and diffusivity. We identified two promising solutions which – together with Sensirion’s current gas meter modules – could be used to build accurate CV sensors that achieve class B or C accuracy. 

The measurement principle is especially promising, which could be implemented as an MEMS Sensor (Figure 6 a) – and thus at a very attractive cost. The achievable accuracy of 2 % would allow class C CV meters to be obtained at low cost.

Figure 6 b shows the combination of Sensirion’s thermal-mass sensor with an optical sensor. This solution would be capable of delivering CV measurement with 1 % accuracy – thus also suitable for class B CV meters. This solution could be more readily acceptable in terms of accuracy, albeit at a higher cost than the MEMS-based solution.

MEMS and optical sensors could be integrated into existing Sensirion gas metering modules to form an energy meter. In this case, OIML R137 class 1.5 flow measurement could be combined with class B or C CV measurement to achieve a simple, cost-effective and scalable solution ready for any number of grid injection points in the future.

Conclusions

Owing to the homogeneous quality of gas in networks to date, the separation of measuring CV (at injection points) and volume (at each consumer) has yielded satisfactory results in the past. With the multiplication of renewable gas injection points, the cost-benefit ratio of scaling the current approach is likely to become unfavorable. While class A and B accuracy are currently used for measuring CV away from the end-point, the technology exists to measure CV closer to the customer at a fraction of the cost. One future solution could be to use direct energy meters based on Sensirion gas metering technology with OIML R140 class B (1 %) or C (2 %) accuracy in CV measurement and class 1.5 accuracy in volume flow measurement. Once the regulations permit direct energy metering with class B or C accuracy in CV measurement, Sensirion will offer an economic solution that is compatible with hydrogen and biomethane to combine volume and CV measurement in one gas meter.

Related sensors