Calculation of Methane and CO2e Emissions from Offshore Flare Systems Currently, for the reporting of CO2 emissions, the flares on offshore oil and gas platforms are assumed to have a combustion efficiency of 98%. The 2% unburnt gas, however, is principally methane. Methane has a Global Warming Potential (GWP) considerably greater than of CO2 (25 over 100 years but 86 over 20 years). Hence, recently there has been an increased focus on the determination of combustion efficiencies and the consequent implications for total CO2e emissions from flares when correctly accounting for the contribution of methane. The University of Alberta (UoA) conducted extensive research into the combustion efficiency of hydrocarbon flares. One of the outcomes of this work was a semi-empirical equation, which expressed the combustion efficiency as function of several variables including: gas exit velocity, wind speed and the Lower Heating Value (LHV) of the un-combusted gas. There is some physical basis for this equation and hence it provides a mechanism to study the effects of the key input parameters on the combustion efficiency and total emissions due to carbon dioxide and the un-combusted methane. This presentation is an analysis of the implications of the semi-empirical flare combustion efficiency (CE) equation, developed by the UoA. Efficiencies of over 98% are achievable at low to average wind speeds. However, due to the exponential nature of the equation, high wind speeds significantly reduce its value. The paper considers the wind speed probability distribution (typically modelled as Weibull) encountered in the North Sea and integrates it with the UoA CE equation to calculate an expected CE value. The simplistic assumption of a single average wind speed leads to an over-estimation of the CE value. The paper explores the use of varying purge rates to the flare to improve CE and mitigate the deleterious impact of wind. Increasing purge rate in the form of fuel gas will increase CO2 emissions due to combustion but by increasing the exit velocity it will reduce CE and hence methane emissions which may thereby reduce the total CO2e from the flare. As an alternative or to complement sampling, the paper explores the use of process simulations to estimate the composition of flare gas and hence its LHV. In particular, the uncertainty in the LHV associated with this approach is assessed.

In the North Sea oil and gas industry, hydrocarbons produced offshore from different sources (e.g. platforms) are often mixed together in shared pipelines as they are transported onshore for further processing. In order to identify who owns the hydrocarbons exiting from the pipeline, equations are required to allocate those hydrocarbons at a component level, e.g. propane, butanes, etc. to the sources they were produced from offshore. The talk: • Describes and discusses pipeline allocation systems using simple examples to illustrate the fairly simple mathematics involved • It focuses on systems that allocate the discharged products at the end of the pipeline (e.g. entry to an onshore terminal) in proportion to the contributions from each of the upstream entry points (e.g. platform export) as though there was instantaneous transfer. • There is no account taken of the transfer time across the pipeline but it is assumed (intuitively) that any differences in the exported versus allocated inlet for each entry point will even themselves out over time. • The presentation illustrates how this assumption may not be correct and therefore the allocation is biased, just due to the form of the allocation equations. • This is demonstrated using mathematical expectation calculations which are explained simply using the returns that might be expected by buying lottery tickets. • Analytical expectation calculations are presented along with the results from Monte Carlo simulations. • Though principally explained using simplified examples, real data is also presented which shows the actual impact of these subtle mathematical bias effects resulting in the mis-allocation of millions of pounds worth of hydrocarbons.

In the North Sea oil and gas industry, hydrocarbons produced offshore from different sources (e.g. platforms) are often mixed together in shared pipelines as they are transported onshore for further processing. In order to identify who owns the hydrocarbons exiting from the pipeline, equations are required to allocate those hydrocarbons at a component level, e.g. propane, butanes, etc. to the sources they were produced from offshore. The talk: • Describes and discusses pipeline allocation systems using simple examples to illustrate the fairly simple mathematics involved • It focuses on systems that allocate the discharged products at the end of the pipeline (e.g. entry to an onshore terminal) in proportion to the contributions from each of the upstream entry points (e.g. platform export) as though there was instantaneous transfer. • There is no account taken of the transfer time across the pipeline but it is assumed (intuitively) that any differences in the exported versus allocated inlet for each entry point will even themselves out over time. • The presentation illustrates how this assumption may not be correct and therefore the allocation is biased, just due to the form of the allocation equations. • This is demonstrated using mathematical expectation calculations which are explained simply using the returns that might be expected by buying lottery tickets. • Analytical expectation calculations are presented along with the results from Monte Carlo simulations. • Though principally explained using simplified examples, real data is also presented which shows the actual impact of these subtle mathematical bias effects resulting in the mis-allocation of millions of pounds worth of hydrocarbons.

A New Orifice Flow Rate Equation The use of three differential pressure (DP) measurements with an orifice meter has been well established for the purposes of meter diagnostics as with the Prognosis validation system. This presentation describes how these multiple DP readings can also be used to develop a new and alternative equation to calculate flow rate, significantly different in form to the traditional flow equation. It utilises momentum balances in the upstream and downstream sections of the orifice and a Bernoulli style energy equation in the upstream section. The resultant equation allows the explicit calculation of the vena contracta diameter and a new velocity head loss coefficient is introduced to account for frictional losses. This head loss coefficient was found to be principally a function of the beta ratio and could therefore be reliably predicted. The new head loss coefficient only accounts for frictional losses in contrast to the discharge coefficient, Cd, of the traditional flow equation, which additionally and principally compensates for the use of the beta ratio of the orifice rather than the ratio of the vena contracta to the pipe diameter. The efficacy of the new equation is illustrated using real data. This new equation provides a potential back-up method to the traditional flow equation in the event of problems with any one of the differential pressure measurements.