Classifications

• • F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
  • F17—STORING OR DISTRIBUTING GASES OR LIQUIDS
  • F17D—PIPE-LINE SYSTEMS; PIPE-LINES
  • F17D3/00—Arrangements for supervising or controlling working operations
  • F17D3/01—Arrangements for supervising or controlling working operations for controlling, signalling, or supervising the conveyance of a product

Definitions

• the present invention relates to a fluid transmission control system and method for controlling the transmission of fluid through a fluid transmission network.
• Natural gas for use as a fuel to provide power to industrial and domestic users via grid networks, turbines, etc., can be obtained from natural gas sources around the world, such as from gas fields or from oil fields (gas that is separated from crude oil at a production facility). Demand for fuel is increasing and often cannot be satisfied by local sources; in such situations gas from a remote source is often required to supplement the quantity and/or quality of a local source. In the situation where there is no local gas source, the demand for fuel is satisfied by gas from a remote source, in particular, in the form of liquefied natural gas (LNG).
• LNG liquefied natural gas
• gas characteristics such as inherent properties (e.g. gas composition) and supply parameters (e.g.
• gas flow rate, temperature, etc. affect the operation or outcome of the process.
• a supplied fuel such as natural gas
• the specifications (required characteristics) of the required gas become more precise, narrowing the range of acceptable gas characteristics and providing a greater variety of characteristic values which must be met for supplies to different users from a certain gas source.
• the specification of a gas may depend on its calorific value or the rate of change of the calorific value (which is preferably low, providing an approximately constant calorific value), which is linked to the "Wobbe Index".
• the Wobbe Index is an indicator of the interchangeability of fuel gases and is frequently defined as a requirement specification of gas supplies and transport utilities.
• the Wobbe Index is used to compare the combustion energy output of different composition fuel gases in an appliance burning such gases.
• users may require a rate of change of calorific value of less than 1%, or a calorific value within a range of ⁇ 5% of a target value. Outside these parameters, factors such as the operational efficiency of the equipment, safety and cost may be affected. Additionally, when gas from a plurality of sources is commingled or processed consecutively, gas composition can vary over time. Operating machinery receiving the gas typically has a limited range within which it can work; that is to say that the tolerance to variations in gas composition can be low. For machinery that is capable of making adjustments to accommodate variations in gas, efficiency is inevitably lost and/or machinery can be damaged by the variation in gas characteristics.
• composition of the gas (that is, the relative amounts of the components of the gas) reaching machinery employed by downstream users is therefore highly important, as the gas energy or calorific value of the gas is dependent upon this composition.
• one or more fuel supplies originating from a remote source such as liquefied natural gas (LNG) supplies
• LNG liquefied natural gas
• these may be used in combination with one another in order to adjust the composition.
• the composition of the fuel can be altered from its original composition by factors such as the quantity and molecular weight of any hydrocarbons extracted during an associated liquefied petroleum gas extraction process, and the amount of any added diluents such as nitrogen gas, which modifies the Wobbe
• the suitability of the gas for downstream users is also dependent upon factors such as the gas export rate (that is, the rate at which downstream users are able to consume the gas), the gas temperature and the gas pressure.
• the present invention provides a system and method for controlling the processing of fluid, such that the fluid supplied downstream of fuel sources to a user meets a predetermined specification.
• a fluid transmission control system arranged to control the transmission of fluid through a fluid transmission network, the fluid transmission network comprising a pipeline having a first end connectable to the output of a fluid processing plant, said fluid processing plant being associated with one or more fluid sources from which fluid is input into the fluid transmission network, and a second end connectable to one or more user systems arranged to use fluid exported by the fluid processing plant
• the control system comprising: measurement means for measuring one or more characteristics of fluid output from one or more of the fluid sources; simulation means for simulating operation of the fluid processing plant in respect of said fluid output from the one or more fluid sources so as to generate data indicative of simulated characteristics of fluid output from the simulated fluid processing plant; prediction means for predicting, based on the simulated fluid output characteristics, characteristics of the fluid corresponding to a time at which the fluid flows through the second end of the pipeline; adjustment
• the process can be changed pre-emptively to coincide with a change, or to compensate for an undesired value, in fuel characteristics, such as fuel composition.
• the process parameters compensate for a change in fuel characteristics such that in some implementations, a product with relatively stable characteristics is exported from a process, despite varying initial fuel characteristics.
• the source fluid input process can be adjusted to account for a measured change in fuel characteristics, for example by adjusting the flow rate, temperature or percentage of fuels from different sources that combine to provide a deliverable gas; the invention is therefore particularly applicable to a gas transmission network having multiple fluid sources, for example a network in which gas from a local gas source is combined with fuel that is stored in a liquid state, and is re-gasified before the combination with the local gas, or a network in which the gas is obtained from a plurality of remote sources owing to the absence of a local gas source.
• Fluctuations or insufficiencies in gas characteristics with respect to a predetermined specification can be predicted and the characteristics can be adjusted accordingly, in order to pre-empt and thus avoid or minimise problems that currently occur upon the delivery of gas having characteristics outside an acceptable range, or which vary over time, to users.
• Such characteristics may include the calorific value, lower heating value (LHV, defined as the amount of heat released by combusting a specified quantity (initially at 25°C or another reference state) and returning the temperature of the combustion products to 15O 0 C), higher heating value (HHV), rate of change of the calorific value, flow rate, temperature, pressure, and Wobbe Index of the gas exported to users.
• LHV lower heating value
• HHV higher heating value
• rate of change of the calorific value flow rate, temperature, pressure, and Wobbe Index of the gas exported to users.
• the hydrocarbon composition of the export gas may be determined (mole% and molecular weight of the various hydrocarbon components).
• the invention allows the management of upstream and downstream conditions to optimise the specification of the delivered gas, which may originate from multiple sources.
• the characteristics of the delivered gas can be controlled by predicting downstream gas characteristics and adjusting the parameters of processes applied to the gas.
• Figure 1 shows a fluid transmission network having a fluid transmission control system according to the present invention
• FIG. 2 is a diagram of the control system of the invention
• Figure 3 a is a flowchart of the processes performed by the system of the invention when measuring fluid characteristics and determining whether a change in characteristics has occurred;
• Figure 3b is a flowchart of the processes performed by the system when applying a fluid processing plant simulation to the network of Figure 1;
• Figure 3 c is a flowchart of the processes performed by the system when applying a predictive pipeline simulation to the network of Figure 1 ;
• Figure 3d is a flowchart of the processes performed by the system when applying a real time pipeline simulation to the network of Figure 1;
• Figure 4 is a flowchart of the process performed by a data manager of the control system according to the invention;
• Figure 5 shows a fluid transmission network having two fluid pipelines and a fluid transmission control system according to the present invention
• Figure 6a is a flowchart showing how time constraints are taken into account by the system of the present invention once a change in a fluid characteristic is measured
• Figure 6b is a flowchart showing the steps taken by the system of the invention in instructing a controller of the system.
• FIG. 7 is a flowchart of an example of the invention in which a characteristic of fluid delivered to the downstream users is maintained at a constant value.
• the network comprises a main pipeline 1 through which fluid, such as natural gas, can flow.
• the main pipeline 1 is connected at one end thereof to the output of a gas processing plant 2, this plant 2 being associated with one or more sources 3 of fuel in the form of a fluid (liquid or gas).
• Another end of the main pipeline 1 is connected to one or more downstream users 4, such as grid networks, industrial users or plants and domestic users.
• the one or more fluid sources 3 may comprise one or more stores or tanks 3 a of fluid such as liquefied natural gas (LNG).
• LNG liquefied natural gas
• LHV lower heating value
• the one or more LNG storage tanks 3 a may be connected to the gas processing plant 2, in which case the gas processing plant 2 comprises a re-gasification facility 6 arranged to convert the LNG from a liquid to a gaseous state.
• the source can comprise fuel in the form of gas.
• LNG tanks 3 a can comprise remotely harvested natural gas which is converted into liquid form before being transported to a location where there is either no local natural gas supply or the local natural gas supply is of insufficient quality and/or quantity, or simply does not meet the specific requirements of the downstream users 4.
• One or more tanks 3a can then be connected to the gas transmission network, via gas processing plant 2, and used to either provide a gas supply to downstream users 4 or to supplement the originally available gas supply.
• sampling, measurement and analysis can be performed on the LNG that is being off-loaded from an LNG carrier to the tank 3 a, the LNG present in the tank 3a or the LNG passing into the gas processing plant 2. Samples may be taken continuously, according to a predetermined schedule, or on demand.
• the gas processing plant 2 typically comprises an extraction facility 7, such as a hydrocarbons or liquefied petroleum gas (LPG) extraction facility, arranged to extract hydrocarbons that are not wanted or required in the export gas leaving the gas processing plant 2.
• an extraction facility 7 is typically arranged to extract propane, butanes and heavier hydrocarbons, thus leaving ethane and methane present in the gas.
• the LPG can advantageously be harvested as a separate fuel product.
• the gas processing plant 2 can also comprise a diluent injection facility 8 for introducing a diluent such as nitrogen into the gas.
• the remaining processed gas is sampled and its characteristics, such as its composition, flow rate and temperature, are measured by one or more further measurement devices 9, before or upon leaving the gas processing plant 2.
• This processed gas is then exported from the gas processing plant 2 and flows along the main pipeline 1 towards the downstream users 4.
• a gas transmission control system will be described in more detail.
• the analysis and extraction processes are controlled by a controller 10, and the measured and/or determined data is stored in a data store 11.
• the controller 10 is operably connected to a server 12, which is typically an Openness, Productivity and Connectivity (OPC) server.
• OPC Openness, Productivity and Connectivity
• the controller 10, data store 11 and server 12 are further operably connected to an upstream Advanced Process Controller (APC) unit 13.
• APC Upstream Advanced Process Controller
• the upstream APC unit 13 can be used to control parameters of the fluid input process that feeds fluid from the source(s) 3 into the gas processing plant 2 and to control parameters of the re- gasification process and/or extraction of LPG, and hence to control characteristics such as the calorific value or LHV of the gas to be exported from the gas processing plant 2, to ensure that the export gas has characteristics such that, when delivered to downstream users, the delivered gas characteristics match pre-set target characteristics or fall within a target range, as described further below.
• the APC unit 13 can increase or decrease the extent of LPG extraction, including by-passing the LPG extraction facility 7, or rejecting LPG back into the gas by adjusting the LPG extraction facility 7 to recover less LPG.
• the upstream APC unit 13 can increase or decrease an amount of nitrogen injected into the gas.
• the upstream APC unit 13 can be configured with any fixed constraints that are required. In this way, the upstream APC unit 13 can control the characteristics of the export gas transported from the plant 2 into the main pipeline 1.
• the upstream APC unit 13 and the downstream APC unit 14 can communicate with one another such that data can be passed between these two APC units 13, 14.
• the controller 10 controls all of the functions of the gas processing plant 2 and is a common component in many modern fluid processing plants.
• the controller 10 communicates with the valve and makes the change. This is achieved by a set of control loops which use an operator defined set point to control the process function or variable (e.g. a valve position).
• the main gas processing plant 2 control is achieved by a large number of control loops which control a variable by using a plant function such as a valve, compressor speed, etc.
• the upstream APC unit 13 determines the most optimum set point for the particular plant function & control loop and then uses the operator interface of the controller 10 to implement this change by writing a new set point into the controller 10 control loop for that function.
• the downstream APC unit 14 is a similar optimisation unit which uses a downstream user process controller to implement any changes to a downstream user process.
• a process management server 20 and a simulation server 30 are provided.
• a process model which is run by the simulation server 30, uses existing gas processing plant 2 conditions or parameters in order to predict the characteristics of gas exported from the plant 2.
• Predictive and real time gas flow models which are also run by the simulation server 30, operate dynamically, and are applied to the gas flowing through the main pipeline 1.
• LNG from one or more LNG tanks 3 a is sampled as it is fed to the gas processing plant 2 in order to obtain real time information on the characteristics, such as flow rate, temperature and composition of the LNG flowing into the processing plant 2.
• LNG flowing into the gas processing plant 2 may be obtained from a single storage tank 3 a or may be a blend of LNG from two or more storage tanks 3a. If the LNG flowing into the processing plant is a blend, the sampling is preferably carried out on the blend.
• the results of the measurement, and any associated analysis such as gas chromatographic analysis, are sent to the process management server 20, which in turn passes the information to a process model run on a plant modelling server 33 of the simulation server 30.
• the process model predicts the properties of the export gas which will be produced by the plant if the existing plant process parameters are used to process the sampled LNG. For example, the online process model can predict the calorific value, the flow rate and the temperature of the gas which would be produced under the existing plant process parameters.
• the predicted data is then fed back to the process management server 20 which determines whether these predicted properties are acceptable by determining whether the prediction falls within pre-set target characteristic ranges which are stored in an information database 22 of the process management server 20. If the predicted characteristics are within the target ranges, then the upstream APC unit 13 makes no adjustment to the gas plant processing parameters. However, if the predicted characteristics do not fall within the target ranges, the process management server 20 instructs the upstream APC unit 13 to make adjustments to the LNG re-gasification process as mentioned above, for example by increasing or decreasing the extent of LPG extraction including by-passing the LPG extraction facility 7, and/or by increasing or decreasing an amount of nitrogen injected into the gas.
• the process model is preferably run ahead of real time and in conjunction with a predictive pipeline model, as described below, to provide a prediction of gas characteristics of the gas flowing through the second end of the main pipeline 1, and on this basis the system can determine whether or not to use the sampled LNG tank(s) 3 a ahead of releasing the fluid from storage and/or processing the fluid. Also, where the fluid that is being processed is a blend of LNG from two or more storage tanks, the system can be used to adjust the relative flow rates of the LNG from the storage tanks.
• the predictive model or "predictive pipeline tracker” runs simultaneously to, and is an exact copy of, the real time model which, as mentioned above, is applied to simulate the gas flowing through the main pipeline 1 ; however, while the real time model simulates the current characteristics of the gas in the pipeline and runs in real time, the predictive model is allowed to run at a faster rate, for example, 30 times faster than the real time model, in order to show the predicted characteristics of the gas, preferably at a predetermined time in the future, such as when the gas flows through the second end of the main pipeline 1.
• the results of the process model are fed into the predictive model.
• the resulting prediction from the predictive model is provided to the process management server 20, which uses the information to determine whether and to what extent the characteristics of the gas to be exported from the gas processing plant 2 should be adjusted such that the export gas meets a required gas specification upon delivery to a user 4; once this determination is made and any necessary adjustment to the processing performed by the gas processing plant 2 is identified by the process management server 20, an instruction in the form of a characteristic setpoint (such as an LHV setpoint) is sent to the APC unit 13 at an appropriate time.
• the APC unit 13 then effects the adjustment at an appropriate time to achieve the required adjustment in exported, and delivered, gas characteristics.
• the APC unit 13 can also be used to switch between different LNG tanks 3a or to adjust the flow rate of LNG from two or more tanks 3 a. Therefore, the predictive model uses the results of the process model to continue the prediction of gas characteristics, such that a prediction of characteristics of the gas passing through the second end of the pipeline, for delivery to end users, can be predicted before any fluid is processed.
• the system can determine whether the fuel from the currently connected local gas source(s) must be supplemented in some way in the first place; that is, whether gas from an additional source, for example in the form of re-gasified LNG as discussed above, is required, and further can determine whether adjustments to the gas processing in the form of an adjustment to the LNG fed into the network or the LPG extraction, bypassing the LPG extraction facility 10, or an injection of nitrogen into the export gas, is required, together with the correct time at which to apply this adjustment.
• gas sampling and analysis is preferably performed at measurement devices 9 before feeding the processed export gas into the main pipeline 1 , in order to determine its precise characteristics.
• the process management server 20 determines whether any of the downstream user's processes need to be altered by the downstream APC unit 14 to account for the measured characteristics of the export gas that will be passed down the main pipeline 1.
• the predictive results of the process model and/or the predictive model can be provided to the downstream APC 14 ahead of real time to allow for such monitoring.
• Downstream users 4 can access the predicted information via a user data interface (not shown) and request adjustments, such as composition and rate adjustments, for example for transition periods.
• the real time model or "live pipeline tracker” uses information regarding the gas characteristics to provide the system, the downstream APC unit 14 and optionally downstream users 4, with information about the gas flowing through the main pipeline 1. This enables the system and downstream users 4 to monitor the gas composition, calorific value, temperature, pressure, and other characteristics of the export gas as described above. If the downstream user's process cannot be altered to account for the predicted or real time gas characteristics, the user can send a request to the process management server 20 for the upstream processes of the gas processing plant 2 or fluid input process to be adjusted so as to provide gas of a required specification. Such a request is typically time- based and may comprise a LHV or flow rate request.
• the process management server 20 receives the request and can instruct the upstream APC unit 13 to adjust the gas processing or fluid input parameters to provide delivered gas in accordance with the requested characteristic for the requested time period.
• the request may include a maximum acceptable characteristic value or change, and in this case the upstream APC unit 13 also notes this as a reference point. Real time measurement and analysis from the measurement devices 9 can ensure that, should the time period be inaccurate, the rate of change of the gas characteristics is altered to ensure that the requested characteristics are met.
• the real time model is also used to determine "residence times" for various sections of the gas transmission network, these residence times being the predicted times for gas to flow from one end to another of the sections.
• the residence time of the main pipeline 1 is the time it takes the export gas to flow from an output of the gas processing plant 2 to the end of the pipeline.
• the system has sufficient time to calculate the required LHV and appropriate action time at which to instruct the APC unit 13, as well as allowing for the reaction time of the APC unit 13 upon receipt of the instruction to be implemented and the residence time of the main pipeline 1, to ensure that the delivered gas characteristics continue to match the pre-set target characteristics or continue to fall within a target range.
• the system is additionally configured to determine adjustment data in relation to the combination of fluid from each of the multiple sources to be fed into the plant 2; for example, the percentage of the total input fluid provided by each fluid source 3 can be varied.
• Such pre-emptive analysis of the fluid can be performed relatively quickly by the system, advantageously minimising gas processing time, effort and cost at the gas processing plant 2.
• the components and operations of the process management server 20 and the simulation server 30 will now be described in more detail.
• fluid characteristics are measured and the data manager 21 determines whether a change in characteristics has occurred, and in figures 3b to 3d, the running of the process model, the predictive model and the real time model, respectively, are described.
• gas characteristic measurement and analysis (SlOl) of the fluid from the fluid source(s) is preferably performed continuously, or alternatively is performed at regular intervals, at measurement devices 5.
• Data in relation to a characteristic, such as the LHV, is routinely passed (S 102) to the controller 10 and is stored or recorded in the data store 11.
• the process management server 20 comprises a data manager 21, which routinely requests, from the controller 10, data on key fluid characteristics stored in the data store 1 1.
• the data manager 21 may request such data at set intervals, for example.
• This data is then passed (S 103) through the OPC server 12 to the data manager 21.
• the data manager 21 compares (S 104) the most recently received value(s) for the LHV stored within the information database 22 of the process management server 20 to assess whether a change has occurred. If the LHV is shown to be unacceptably lower (or higher, as appropriate) than a predetermined threshold value, such as a data tolerance value, which is stored in the database 22, further action by the data manager 21 is required. Referring to Figure 3b, the data manager 21 sends (S201) the new LHV, together with all other measured gas characteristics (which may not have changed in conjunction with the LHV) to a process server 33 of the simulation server 30, and requests a list of critical information that is required by the server in order to run the process model.
• a predetermined threshold value such as a data tolerance value
• the process server 33 feeds (S202) the new LHV, together with all other measured gas characteristics, into the process model.
• the process model runs (S203) ahead of real time as described above, in order to determine information relating to predicted characteristics of the export gas produced by the gas processing plant 2 based on the current plant 2 processing parameters. This information is then sent via the process server 33 and the data manager 21 to the information database 22 (S204), where it is recorded.
• Such information is part of the requested critical information requested by the data manager 21 , which may include information relating to time, pressure, temperature, LHV, gas flow rate, etc.
• the data manager 21 sends (S301) the predicted gas characteristics output by the process model, which has been run using the new measured LHV and all other measured gas characteristics (which may not have changed in conjunction with the LHV), to a predictive server 31 of the simulation server 30, and requests a list of critical information that is required by the server in order to run the predictive pipeline model.
• the predictive server 31 feeds (S302) the results of the process model into the predictive pipeline model.
• the predictive pipeline model runs (S303) ahead of real time as described above, in order to determine information relating to predicted characteristics of the gas when it flows through the second end of the main pipeline 1, having been processed according to the gas processing plant 2 parameters used in the process model. This information is then sent to the predictive server 31.
• Such information is part of the requested critical information requested by the data manager 21, which may include information relating to time, pressure, temperature, LHV, gas flow rate, etc.
• the critical information is then sent (S304) to the data manager 21.
• the data manager 21 comprises a custom calculation package 23, and uses this to determine (S305) the type and extent of an adjustment required by the system to counteract the anticipated characteristic value of the gas flowing through the second end of the pipeline that has been predicted by the predictive model.
• S305 the type and extent of an adjustment required by the system to counteract the anticipated characteristic value of the gas flowing through the second end of the pipeline that has been predicted by the predictive model.
• the time and LHV may be used to calculate what adjustment is required by the gas processing plant 2 processes, or the fluid input process, and the exact time at which this adjustment should be implemented.
• Such information is determined as setpoints sent as instructions to the upstream APC unit 13, and is described in more detail below in relation to Figure 4.
• step S305 the results of step S305 are recorded (S306) in the information database 22 for a "patience time" period, so that this data can be checked or confirmed.
• the patience time can be defined as a period of time for which the upstream APC unit 13 waits whilst the data manager 21 confirms (S307) that any adjustment data determined by the data manager 21 is consistent, and hence may be acted upon.
• the data manager 21 receives a subsequent set of data sent, as in step S304, from the predictive server 31, and compares these data against the adjustment data stored in the information database 22, typically employing an algorithm in order to assess whether the adjustment data is "true” or "false” data.
• This process can be performed with a number of sets of data such that any number of sets of data are compared in providing the confirmation.
• Each of the predictive model runs is based upon the results of the process model, which in turn are based on measured characteristics of the fluid from the one or more fluid sources 3.
• the adjustment data are transferred (S308) into an action list of setpoints for the upstream APC unit 13 stored within the information database 22, for instructing the upstream APC unit 13 at an appropriate time to allow correct implementation of the setpoint.
• Each of the setpoints in the action list is time stamped with this appropriate instruction time.
• the action is discarded by the data manager 21.
• the implementation or action time is calculated for fluid that has already left the source 3, or for which a predetermined release time is known.
• the action time can be based on an unknown release time (for example, if the time at which a new storage tank 3 a must be used depends on the time at which a previous tank 3 a is emptied, which in turn depends on actual processing and flow rate), in which case a flag is sent to the data manager 21 upon starting use of the new storage tank 3a to calculate the exact action time.
• the data manager 21 sends (S 309) the new setpoint for the LHV to the APC unit 13 via the OPC server 12, and the adjustment is implemented (S310) by the controller 10. Steps S301 to S310 are repeated every time a change in excess of a predetermined threshold, to any of the gas characteristics, is determined.
• the real time modelling is preferably performed in parallel with the predictive modelling.
• the real time model simulates the current characteristics of the gas in the main pipeline 1 in real time, while the predictive model runs at a faster rate, ahead of the real time model, in order to determine the predicted characteristics of the gas at a time in the future.
• the data manager 21 sends (S401) the predicted gas characteristics output by the process model, which has been run using the new measured LHV and all other measured gas characteristics (which may not have changed in conjunction with the LHV), to a real time server 32 of the simulation server 30, and requests a list of critical information that is required by the server in order to run the real time model.
• the real time server 32 feeds (S402) the results of the process model into the real time model.
• the real time model runs (S403) as described above, using a characteristic tracking software element to time stamp the characteristic data.
• the time stamped characteristic data is then sent via the real time server 32 and the data manager 21 to the information database 22 (S404).
• Such information corresponds to the critical information requested by the data manager 21, which may include information relating to time, pressure, temperature, LHV, gas flow rate, etc.
• the information database 22 records the time and characteristic information received.
• the information database 22 can be accessed by system operators and engineers, as well as by downstream users 4 via the internet using the user data interface (not shown) to view the time and gas characteristic information as it is tracked down the pipeline.
• the process management server 20 employs suitable software, while methods for configuring a simulation model using the HYSYS application are known to those skilled in the art and can be employed by the simulation server 30 and associated predictive server 31, real time server 32 and process server 33.
• figure 4 shows a flowchart of the process performed by the data manager 21 in calculating the adjustment data, in terms of a time and a LHV (the LHV characteristic can be replaced by any characteristic that is controllable with respect to time).
• the data manager 21 preferably runs through the process of Figure 4 continuously; in overview, the process involves the data manager 21 running the predictive model ahead of time and recording the results.
• the data manager 21 then uses the custom calculations package 23 to work out what to adjust and when, based on the results of the predictive model.
• the adjustment data is then stored for the patience time before being sent as a setpoint to the upstream APC unit 13. Each time a change in a measured characteristic, and hence a change in the results of the process model, occurs, the same predictive actions will occur.
• An appropriate algorithm in relation to the characteristic will be applied as part of the custom calculations package 23 for the data manager 21.
• any change to the gas characteristics determined as a result of the process model, and real time data are received by the data manager 21 and are stored in the information database 22.
• the data manager receives the predicted characteristic(s) and critical information from the predictive server 31 (as in step S304 of Figure 3c).
• the data manager 21 checks (S5O3) whether the predicted characteristic and critical information meet a value (for example, a threshold value), or are within a target range of values, in relation to current target delivery conditions. If this is the case, no action is taken (S504).
• a characteristic adjustment calculation is requested (S505), which involves the input of critical information from the information database 22 and the implementation of a calculation from the custom calculation package 23.
• a new required LHV value one which is required to ensure that the gas flowing through the end of the main pipeline 1 complies with the target value or range, is calculated at step S506; effecting this change to LHV value necessitates an adjustment to the re-gasification plant process or the fluid input process.
• the LPG extraction process, nitrogen injection level or fluid input parameters are modified accordingly.
• the custom calculation package 23 is again employed to provide an action time calculation at which the adjustment must be instructed to the upstream APC unit 13.
• This action time is calculated at step S508, and again critical and/or characteristic data may be accessed from the information database 22 for the calculation.
• the adjustment data is timestamped with the action time at which the instruction is to be sent to the upstream APC unit 13, and at step S510 the timestamped adjustment data is stored in the upstream APC unit 13 action list in the information database 22.
• a predictive model for a supply pipeline (not shown) linking the fluid source(s) 3 and the gas processing plant 2 may be provided.
• the results of this initial predictive model (“final predictions") may in this case provide characteristics input data for the process model, the results of which in turn provide input data for the predictive model running on the pipeline 1 ; these final predictions are then used in the process of Figure 4, as described above.
• a full prediction of the characteristics of the delivered gas can be performed before any fluid is released; furthermore, the characteristics of the input fluid to be released, or the gas processing plant 2 parameters, can be adjusted as discussed above on this basis.
• re-timestamping of the action time at which the upstream APC unit 13 is to be instructed with an adjustment to the gas processing plant 2 parameters, based on the residence time of the additional supply pipeline may be necessary.
• a suitable procedure for such re-timestamping is discussed below with reference to steps S511 and S512 of Figure 4, and steps S606 to S608 of Figure 6b.
• the current timestamp parameters are checked (S511) against those used previously in calculating the action time at step S508, and an assessment is made (S 512) as to whether or not the timestamp parameters have changed. For example, if the fluid originating at the fluid sources 3 and measured at the measurement devices 5 shows a decrease in LHV, the custom calculation package 23, optionally in association with the predictive model, calculates or predicts the required LHV of the input fluid together with the time at which the gas processing plant 2 will have to make an adjustment to compensate for the decrease in LHV, in accordance with steps S505 and S507 above.
• the rate at which this input fluid reaches the gas processing plant 2 may, in some embodiments, be determined by the flow rate of the gas through the supply pipeline linking the fluid sources 3 and gas processing plant 2. As mentioned above, this residence time is the time taken for that gas to flow from one end (the fluid sources 3) to another (the input of the gas processing plant 2). Should the gas flow rate into the supply pipeline decrease at some point in the future, then the gas with the decreased LHV is already in the supply pipeline; however, the time it will take to move along the supply pipeline will change, that is, the residence time of the supply pipeline will increase.
• the upstream APC unit 13 will therefore need to make the same change to the gas processing but at a different time, hence there is a requirement to re-run and re-timestamp the setpoint data; this process will be described further in connection with Figure 6b below.
• step S 507 if the timestamp parameters have changed, the process returns to step S 507 and the action time calculation is performed again with the changed timestamp parameters.
• This feedback loop is provided to ensure that the upstream APC unit 13 does not make changes to the gas processing plant 2 or fluid input processes solely based on the change in gas characteristics. If it is assessed at step S512 that the timestamp parameters have not changed, the process progresses to step S513 where an assessment is made as to whether or not the action time of the timestamp is equal to the actual current time. If the action time has not been reached, the data manager 21 does not send the instruction or setpoint but waits (S 514) and returns to perform steps S51 1 onwards again.
• the setpoint is sent by the data manager 21 and executed by the upstream APC unit 13 at step S515, so that the calculated adjustment is implemented at the calculated action time.
• the predictive model has additional functions when run on the main pipeline 1; one is to act as a verification to ensure that the custom calculation package 23 has made the correct adjustments, and the other applies in more complex calculations (for example, when multiples are changing), when the predictive model can be run to help the custom calculations package 23 to achieve the correct change.
• the predictive model can provide more resolution to the calculations within the custom calculation package 23.
• Figure 5 shows an embodiment in which the gas transmission network comprises an associated pipeline 50, connected at one end to a fluid source 51, and at another end to the main pipeline 1, forming a junction 52 at which the two pipelines connect, such that fluid flowing through the pipelines combines at the junction 52 and continues to flow downstream thereof to the downstream users 4.
• the fluid source for the associated pipeline 50 may take the form of one or more incoming or upstream gas fields and/or associated gas from one or more incoming or upstream oil fields.
• associated gas is meant gas that is separated from produced oil and water at a production facility.
• One or more characteristics of the natural gas output from the upstream gas fields and/or oil fields 51 are measurable at one or more measurement devices 53, which preferably include composition analysis equipment, such as gas chromatographic analysis equipment that allows identification of the components of the gas and the relative amounts thereof.
• the measurement devices 53 are preferably situated in the vicinity of a metering station 54. Samples of the gas may be taken from the gas at an appropriate point as it enters or flows through the associated pipeline 50 from the upstream gas fields and/or oil fields 51, and such samples may be taken continuously, according to a predetermined schedule, or on demand.
• the predictive and real time models are run for the associated pipeline 50, and are considered by the control system alongside the results of the process model, predictive model and real time model run on the main pipeline 1 , in order to predict and necessary adjustment to the fluid input process, or the gas processing plant 2 processes, such that the combined gas reaching a downstream user 4 inlet meets a required specification.
• the predictive model is again allowed to run at a faster rate than the real time model, in order to show the predicted characteristics of the gas as it reaches the junction 52.
• Such a prediction is provided to the process management server 20, which uses the information to determine whether and to what extent the characteristics of the fluid to be input into the network from a plurality of fluid sources 3, such as LNG storage tanks 3a, and/or the gas exported from the gas processing plant 2, should be adjusted such that when the export gas is mixed with the associated gas, the combined gas meets the required gas specification upon delivery to the user; once this determination is made and any necessary adjustment to the processing performed by the gas processing plant 2 is identified by the process management server 20, an instruction in the form of a characteristic setpoint (such as an LHV setpoint) is sent to the upstream APC unit 13 at an appropriate time. The APC unit 13 then effects the adjustment at an appropriate time to achieve the required adjustment in exported, and delivered, gas characteristics.
• downstream users 4 can access the predicted information via the user data interface 40 and request adjustments, such as composition and rate adjustments, for example for transition periods.
• the system can therefore determine whether the gas from the upstream gas fields and/or oil fields 51 must be supplemented in some way in the first place; that is, whether gas from an additional source, for example in the form of re-gasified LNG as discussed above, is required, and further can determine whether adjustments to the re-gasification process in the form of an adjustment to the LPG extraction, bypassing the LPG extraction facility 7, or an injection of nitrogen into the export gas, is required, together with the correct time at which to apply this adjustment.
• the system can additionally of alternatively determine the extent to which the analysed fluid from the LNG tanks 3a should be adjusted before inputting the fluid into the fluid transmission network, for example by determining the temperature or flow rate of the fluid, or by determining the amount of fluid from each of a number of LNG tanks 3 a which must be mixed to provide a particular input fluid specification.
• the real time model or live pipeline tracker uses information regarding the associated gas characteristics to provide the system, and optionally downstream users 4, with information about the associated gas flowing through the second pipeline 2.
• a downstream APC unit (not shown) similar to that described with respect to Figure 1 may also be provided. This enables the system and users 5 to monitor the gas composition, calorific value, temperature, pressure, and other characteristics of the gas in the main pipeline 1 and the associated pipeline 50, as well as that of the combined gas flowing downstream of the tie-in j unction 52.
• the residence time of the associated pipeline 50 is longer than the residence time of the main pipeline 1 such that the gas that enters the associated pipeline 50 takes longer to reach the junction 52 than the gas flowing through the main pipeline 1.
• the difference in residence time should be great enough to ensure that the system has sufficient time to calculate the required LHV and the appropriate action time at which to instruct the APC unit 13.
• the difference in residence time should also be great enough to allow for the reaction time of the APC unit 13 upon receipt of the instruction to be implemented and to allow for the residence time of the main pipeline 1. Accordingly, the difference in residence time is sufficiently great enough to ensure that the combined gas characteristics continue to match the pre-set target characteristics or continue to fall within a target range.
• the residence time of any additional network sections, such as an initial pipeline (not shown) connecting the LNG tanks 3a to the gas processing plant 2, must also be taken into account.
• the predictive and real time models are run on the associated pipeline 50 as described in Figures 3c and 3d above, and the results of these models are used by the data manager in addition to those obtained by running the models on the main pipeline 1 , in performing the processes and calculations of Figure 4. As mentioned previously, such steps can be performed in order to predict the characteristics of the fluid input from the LNG tanks 3a.
• the custom.calculation package 23 calculates or predicts the required LHV of the export gas together with the time at which the gas processing plant 2 or fluid input process will have to make an adjustment to compensate for the decrease in LHV, in accordance with steps S505 and S507.
• the rate at which this associated gas reaches the tie-in junction 52 is determined by the flow rate of the gas through the associated pipeline 50. As mentioned above, this residence time is the time taken for that gas to flow from one end (the upstream gas fields and/or oil fields 51) to another (the tie-injunction 52).
• the system can be programmed to evaluate a predetermined number of measurements and to run the process model and then the predictive model for each of these before allowing an adjustment to be instructed. For example, ten (eight, five or three) samples can be taken, analysed and used in the process model, and the results used in the predictive model, before the system will confirm the adjustment and action time. As additional verifications are made, the error margin of the adjustment data is reduced. If, when performing a verification with a subsequent set of data, it transpires that no action is required (for example, in the case that the previous data was erroneous), the data can be re- timestamped or the setpoint removed, as appropriate.
• the custom calculations package 23 of the predictive model takes a number of factors into account when calculating predicted gas characteristics for the gas flowing to the downstream users.
• a "LHV balance" calculation is performed as described above to calculate the required LHV (or other required gas characteristic) to be exported by the gas processing plant 2, in order that, when mixed with the associated gas, the combined gas flowing downstream of the tie-injunction 52, and therefore delivered to a downstream user 4, has a specific LHV (or other required gas characteristic).
• LHV balance calculation can calculate the required LHV of input fluid for the gas processing plant 2 to be input by mixing fluid from multiple fluid sources 3.
• a "flow balance" calculation is performed. This calculates the Standard Volumetric Flowrate of gas to be exported by the gas processing plant 2 in order that, when mixed with the associated gas, the combined gas flowing downstream of the tie-injunction 52, and therefore delivered to a downstream user 4, has a specific flow rate.
• a LHV balance calculation can calculate the required LHV of input fluid to be input from one or more fluid sources 3.
• a further LHV balance calculation can be performed to aid the prediction of the rate of change of the LHV of the gas that is delivered to a downstream user 4.
• a "residence time” is calculated (typically by the real time model) for each section of the gas transmission network.
• these residence times are the predicted times for gas to flow from the export of the gas from the gas processing plant 2 to the tie-in junction 52; from the associated gas source metering station 54 (or wherever the measurement devices 6 are situated) to the tie- in junction 52; and from the tie-injunction 52 to a downstream user 4 inlet.
• the residence times allow the system to consider how many times the process model and the predictive model can be run (factoring in the patience time) to verify the adjustment data and hence reduce the error margin.
• Figure 6a shows how time constraints are taken into account by the system once a change in the LHV is determined by the data manager 21.
• the system determines whether the residence time of the associated pipeline 50 (P2RT) is greater than the residence time of the main pipeline 1 together with any additional residence time of the main section of the gas transmission network, such as the residence time of a supply pipeline linking the fluid sources 3 and input of the gas processing plant 2, (PlRT) of the system, plus any additional time that must be factored in by the system, such as a reaction time of the upstream APC unit 13 (ART) and a safety margin time (SMT).
• P2RT residence time of the associated pipeline 50
• ART reaction time of the upstream APC unit 13
• SMT safety margin time
• the safety margin time includes the time that it takes the system to calculate the required LHV of the export gas output from the gas processing plant 2 based on the results of the predictive model. If this is the case, there is sufficient time for the system to calculate, instruct, react to and implement the adjustment and for the gas to flow along the main pipeline 1 to the tie-injunction 3, and the process moves to step S602. If there is insufficient time to perform these actions, no action is taken and the process is repeated when a further change to the gas characteristics is determined.
• step S602 the system starts a run of the process model, the results of this being fed into the predictive model.
• step S603 the system determines whether a sum of the process model simulation time plus the predictive model simulation time (PMST) is greater than the residence time of the main pipeline 1 (and any additional residence times) (PlRT) of the system plus the reaction time of the APC unit 13 (ART) and the safety margin time (SMT). If this is not the case, the process returns to step S 602 and the process model is re-started; alternatively, if this is the case, this step ensures that the results of the predictive model will allow an adjustment at an appropriate time to counteract the determined change in the characteristic of the associated gas.
• PMST predictive model simulation time
• PlRT reaction time of the APC unit 13
• SMT safety margin time
• step S604 the predictive model is stopped and the calculated characteristic adjustment, for example in the form of a required LHV of the export gas or input fluid, is read.
• PET predictive model simulated end time
• OTS PET - PlRT - ART
• the data manager 21 reads the action data from the action table of the information database 22.
• the action table may include a flag field indicating the status of an action to be "true” if it has already been written to the APC unit 13 but is still stored in the database 22, or "false” if the action is yet to be written to the APC unit 13; in this case the data is only read if the status flag is false.
• the current residence time of the main pipeline 1 (and any additional residence times) (ClRT) is identified by running the real time model on the main pipeline 1 (and any additional sections of the network, as required) in step S607.
• a "reconciled timestamp” is calculated, by subtracting the difference between the residence time calculated during the process of Figure 6a and the current residence time, from the original timestamp calculated in the process of Figure 6a: RTS - OTS - (PlRT - ClRT).
• the reconciled timestamp is then stored in a relevant field in the action table.
• This reconciled timestamp is the new time at which the adjustment data needs to be written or instructed to the APC unit 13, owing to the change in the residence time of the main pipeline 1 (and any additional residence times) from the time at which the predictive run took place, to the current time; such a change could occur due to a change in the flow rate of the gas.
• the data manager 21 determines whether there is a reconciled timestamp entry stored in the action table that is within a certain time, for example within one minute, of the current time. If there is such a reconciled timestamp entry, the corresponding adjustment data, such as a new LHV, is written (S610) to the APC unit 13 and the entry is marked as "true” accordingly. If no such entry is present in the action table, the system waits, at step S611 , for an appropriate predetermined time, such as 30 seconds, and then returns to the start of the process.
• an appropriate predetermined time such as 30 seconds
• the adjustment is instructed as a single instruction of a specified rate of change, rather than instructing the APC 13 with a plurality of staged or stepped discrete value changes, which results in the desired overall adjustment, as typical gas processing plants 2 deal with a "rate of change" adjustment more effectively; however, both implementations are possible.
• FIG. 7 shows a flowchart of a simplified example in which a characteristic of gas delivered to the downstream users 4 is maintained at a constant value.
• step S701 a change in the LHV value of the associated gas measured by the measuring devices 53 is determined by the data manager 21.
• This data is sent to the predictive server 31 and the predictive model is run (S702) with this new LHV value as an input, and as a result LHV of the gas in the associated pipeline 50 is predicted to reduce in step S703.
• the process model and predictive model are then run (S 704) on the main pipeline 1.
• the results of all models are sent to the data manager 21, which determines, in step S 705, the type, extent and action time of one or more adjustment(s) required to the exported gas from the gas processing plant 2, and/or to the input of gas from fluid sources 3 into the network, in order to counteract the reduction in the LHV of the gas in the associated pipeline 50 when the pipelines meet at the tie-injunction 52.
• the APC unit 13 is instructed to implement the adjustment (S706) at the appropriate action time, and may implement the instruction after an associated patience time. In such a case, any sharp variation in the LHV of the gas delivered at a downstream user's pipeline inlet is avoided, and a steady LHV is maintained.
• the gas transmission control system may be implemented where gas is supplied to multiple users situated downstream of the junction 52 at which the main pipeline 1 and the associated pipeline 50 converge.
• the transmission network may supply three or four power stations, all of which have specific gas characteristic requirements, and one of which has a narrower acceptable range for the LHV of rate of change of the LHV than the other stations.
• gas having characteristics that meet the narrower specification, and which therefore meets all of the stations' specifications can be produced. Adjustments can be made to the gas characteristics depending on which power stations are on or require a supply at a certain time, in order to manage gas transmission network and control system efficiency and costs.
• each tank 3 a may contain LNG originating from a different remote source (for example, LNG originating in different countries) having different characteristics, so the fluid characteristics may be measured and analysed when transferred to the tanks 3 a for storage, so that the system is aware of the characteristics of the LNG in each tank 3a and can use the supplies accordingly.
• a different remote source for example, LNG originating in different countries
• Such use can be predicted in advance of supplying the fluid as described above. Further measurement and analysis can then be performed upon exporting LNG from a particular tank 3a to the gas processing plant 2, as a confirmation of the characteristics or to detect any changes, for example due to contamination, that may have occurred during storage.
• LNG from multiple original sources may be stored in each of the LNG tanks 3a.
• the characteristics of the LNG can be measured and analysed when passing from the tank 3a to the gas processing plant 2.
• the gas characteristics of each are measured and analysed whilst remaining in storage, such that the results thereof can be considered by the system and the LNG can be released and, if necessary, mixed in appropriate quantities to minimise alterations to the composition of the mixed LNG stream that is fed to the gas processing plant 2.
• Such pre-emptive analysis of the LNG can be performed relatively quickly by the system, advantageously minimising gas processing time, effort and cost at the gas processing plant 2.
• the invention provides significant advantages for downstream gas users in reliability, maintainability and efficiency of combustion.
• gas fired turbine power stations are particularly sensitive to gas quality changes, particularly in LHV and Wobbe Index. The range of both of these characteristics is required to be within a defined range of values to ensure acceptable operation, while failure to adhere to this can lead to poor power station availability.

Landscapes

• Engineering & Computer Science (AREA)
• Mechanical Engineering (AREA)
• General Engineering & Computer Science (AREA)
• Management, Administration, Business Operations System, And Electronic Commerce (AREA)
• Flow Control (AREA)

Applications Claiming Priority (2)

Publications (1)

Family

ID=40343881

Family Applications (1)

Country Status (3)

Families Citing this family (5)

Family Cites Families (8)

• 2008
  • 2008-12-18 GB GB0823171A patent/GB0823171D0/en not_active Ceased
• 2009
  • 2009-12-16 EP EP09796029A patent/EP2368067A2/de not_active Withdrawn
  • 2009-12-16 WO PCT/GB2009/002905 patent/WO2010070288A2/en not_active Ceased

Non-Patent Citations (1)

Also Published As

Similar Documents

Legal Events