Un-accounted for gas (UAFG) is the difference between the total measurements of gas injected into a pipeline system and the total measurements of gas withdrawn from the same system with a correction for any changes in the quantity of gas stored in the system (linepack) over the measurement period. High pressure gas transmission pipelines do not normally leak to any material extent and yet it is normal for transmission pipelines to have a non-zero UAFG. This arises because of unknown biases in both the injection (input) and withdrawal gas flow measurements.
The measurement of UAFG is complicated by any errors in the measurement or estimation of linepack at the start and end of the chosen measurement period. Because of this, day-by-day UAFG determination tends to give a wide scatter in values which may not be indicative of the actual losses/measurement biases for that day as shown in figure 1. Thus, relatively long measurement periods may be required to get a reliable indication of the average UAFG.
Although gas transmission systems do not normally leak, transmission UAFG can sometimes be caused by actual losses from the system. Losses from the system may not be leakage to atmosphere but may be un-metered flows such as through meter by-passes into other pipeline systems.
Undetected un-metered offtakes contribute to UAFG and this can become especially challenging where an offtake can either inject into or withdraw gas from the transmission system under different flow conditions. It is possible in this situation for gas to flow un-metered into the transmission system from a slightly higher pressure distribution system and thus effectively conceal the existence of other potential losses from the system.
Flow and linepack measurement biases may change with time, as they may be flow rate, pressure, temperature, meter performance, gas composition or heating value related effects. A zero UAFG does not necessarily indicate perfect measurement without any losses as measurement biases and loss effects may cancel.
The measured annual average UAFG in the VENCorp operated transmission system in Victoria is approximately minus 0.25 per cent for the past two years as shown in figure 2. The minus sign indicates that the withdrawal measurements exceed the injection measurements. This is the equivalent of “˜making’ gas in the pipeline in that the measurements of gas flowing out of the transmission system exceed the measurements of gas flowing into the transmission system. This average 0.25 per cent level of agreement between the injected and withdrawn gas flows is very good considering the total uncertainty of both the injection and withdrawal energy measurements exceed ±1 per cent. Nonetheless, – 0.25 per cent UAFG in a transmission system moving up to 3 or 4 million dollars (wholesale) worth of gas a day is significant financially.
As figure 1 shows, the measured day-to-day variations in un-accounted for gas in Victoria sometimes exceed ±1 per cent of the total throughput but the relative contribution of the uncertainty of the linepack measurements and biases in the metering are unknown. Investigations of flows on days with unusually large values of UAFG usually do not yield any specific reason for the large value.
The biases in the injection and withdrawal measurements are included in the long term un-accounted for gas. These biases are unknown and may be positive or negative and may vary from meter to meter and may depend on total gas conveyed and facilities involved. The relative contribution to UAFG from actual losses or gains or from metering biases is unknown.
Market Settlement
In the settlement of the Victorian gas market, UAFG is allowed for in the Victorian Gas Industry Market and System Operations Rules (MSO Rules). Where the “˜actual withdrawals’ are related to the “˜measured withdrawals’ thus:
- AWT = MWT / (1 – UAFGT)
Where:
- AWT is the adjusted quantity of gas that is treated as having been withdrawn by a Market Participant from the transmission system in a scheduling interval;
MWT is the measured quantity of gas withdrawn by a Market Participant from the transmission system in a scheduling interval; and,
UAFGT is the transmission UAFG and is set to 0.0% in the MSO Rules.
The MSO Rules have specified values for both the transmission and distribution un-accounted for gas. The transmission un-accounted for gas is set to 0 per cent, presumably on the basis that no losses are expected.
On any given day the metered injections and withdrawals depend on:
- Injections scheduled by the market operator to meet:
- demand forecasts;
- controllable withdrawals; and,
- end-of-day linepack targets.
- Differences between scheduled and actual flows.
- Any losses/leakage/un-metered offtakes.
- Unknown meter biases.
The start-of-day schedule for gas injections is based in part on market participant forecast withdrawals. Any forecast or delivery error (under or over delivery or withdrawal) is allowed for by changes to the schedule at each scheduling period during the day. In Victoria, the gas flow schedules are normally updated as required every four hours, except for an eight hour scheduling period over night.
In the Victorian gas market settlement, the difference between the metered injections and withdrawals is cleared through imbalance payments, deviation payments and the market “˜linepack account’. The linepack account provides the mechanism whereby the total financial impact of over or under injections or withdrawals not covered by imbalance or deviation payments can be reconciled on each day. Market participants who are net withdrawers effectively fund those who are net suppliers. The impacts of forecast errors and the differences between actual and scheduled flows are priced differently in the linepack account.
The linepack account can be either positive or negative on any given day and the total amount is divided amongst the market participants in proportion to the quantity of gas withdrawn by each participant. As the linepack account is based on metered injections and withdrawals, it “˜accounts for’ the UAFG even though the notional UAFG is set to 0.0 per cent by the MSO Rules. This accounting for the UAFG is in monetary terms rather than energy (or volume) terms. The relationship between the energy UAFG and the monetary UAFG is complicated because the forecast withdrawal and delivery quantity error components of the linepack account are priced differently.
Who Pays for UAFG?
The UAFG cost that a market participant bears on any given day will be proportional to their (withdrawal) market share. The Victorian gas market currently operates with a negative UAFG and thus the market participants receive payments that compensate for the fact that measurements indicate that more gas is withdrawn from the transmission system than is injected into it. The financial impact of UAFG depends on the value of gas on the day, compliance with injection schedules and the accuracy of withdrawal forecasts. A long-term average zero un-accounted for gas in energy terms may still have a non-zero net financial impact because of day-to-day and intra-day fluctuations in gas price and UAFG.
If there is no forecast error or injection quantity error by any participant, the linepack account would represent just the UAFG. This effect comes about because the linepack account is based on the difference between injection and withdrawal measurements on each day and these measurements include the unknowable biases and losses (or gains) that contribute to the overall UAFG. In reality, forecast and injection quantity errors are common and thus the linepack account contains significant non-UAFG components.
Theoretically, if a market participant could arrange to match their injections and withdrawals on each day to exactly match their share of the UAFG for that day, they could reduce their impact on the UAFG component of the linepack account to zero, but would still be liable for a share of the overall linepack account. In practical terms this would not be feasible anyway as most withdrawals from the system are uncontrolled, and although the long term UAFG is fairly constant, the day-to-day variations are unknowable because of the difficulty in determining accurately the actual linepack at the start and end of each day. The actual UAFG that can be assigned to an individual market participant is unknowable. The other issue is that forecast errors and injection quantity errors by other participants will be added to the linepack account, which will then be allocated across all the participants. A full discussion on the financial settlement of the Victorian market is beyond the scope of this article.
Conclusion
The market participants pay for UAFG based on a daily measure of the difference between the measurements of injections and withdrawals. This measurement includes the change in linepack, any gains/losses from the system, and any biases in the metering and energy calculation. The actual magnitude of the payment varies in proportion to the withdrawals by the market participant but is also dependent on day-to-day and intra-day fluctuations in gas price, any forecasting or injection quantity errors and the unknowable day-to-day fluctuations in actual UAFG. The impact on participants of forecast errors and injection quantity errors are greater than and effectively “˜hide’ the UAFG impact.
Acknowledgements
The author wishes to acknowledge the valuable advice from Roger Shaw of VENCorp in the preparation of this paper. The author also wishes to thank VENCorp for permission to publish this paper.
Editor’s note: This article, detailing the method that VENCorp runs its business, does not necessarily reflect the way other pipeline operators work. VENCorp is the market operator in Victoria and practices in other states and operations may vary.