Control Of A Thermal Cyclic Process

Abstract

The invention relates to a method for controlling a thermal cyclic process, in particular an Organic Rankine Cycle (ORC), which is operated with a working medium in conjunction with a dynamic heat source, whereby the method comprises the following steps: (a) determination of a setpoint value of a process variable of the thermal cyclic process from a value of an input parameter or respective values of a plurality of input parameters of the thermal cyclic process; (b) control of the thermal cyclic process with the determined setpoint value of the process variable as a target variable of the control; and (c) repeated execution of steps (a) and (b) when at least one value of the input parameters changes.

Claims

1. Method for controlling a thermal cyclic process, in particular an Organic Rankine Cycle (ORC), which is operated with a working medium in conjunction with a dynamic heat source, comprising: (a) determination of a setpoint value of a process variable of the thermal cyclic process from a value of an input parameter or respective values of a plurality of input parameters of the thermal cyclic process; (b) control of the thermal cyclic process with the determined setpoint value of the process variable as a target variable of the control; and (c) repeated execution of steps (a) and (b) when at least one value of the input parameters changes.

2. Method according to claim 1 wherein the determination of the setpoint value of the process variable of the thermal cyclic process comprises a calculation of the setpoint value from a predetermined function, particularly from a polynomial function, into which the input parameters are input as variables, or wherein the determination of the setpoint value of the process variable of the thermal cyclic process comprises a reading of the setpoint value from a predetermined table depending on a value of an input parameter or on respective values of a plurality of input parameters of the thermal cyclic process, wherein preferably there is interpolation or extrapolation between the table values.

3. Method according to claim 2 with, carried out before the step (a), the additional step: designation of the function for calculating the setpoint value of the process variable by means of carrying out trials and/or from model equations, or designation of the table for determining the setpoint value of the process variable by means of carrying out trials and/or from model equations.

4. Method according to claim 3 wherein the designation of the function for calculating the setpoint value of the process variable comprises a maximisation of the product of an efficiency of the heat transfer from the heat source to the cyclic process and an efficiency of the cyclic process, or wherein the designation of the table for determining the setpoint value of the process variable comprises a maximisation of the product of an efficiency of the heat transfer from the heat source to the cyclic process and an efficiency of the cyclic process.

5. Method according to claim 1 wherein at least one further process variable is controlled in a corresponding manner.

6. Method according to claim 1 wherein the process variable is or the process variables are the vaporisation temperature or the vaporisation pressure of the working medium and/or the condensation temperature or the condensation pressure in a condenser of the thermal cyclic process and/or the vapour temperature or the vapour pressure at the outlet of an expansion machine of the thermal cyclic process.

7. Method according to claim 1 wherein the input parameter comprises or the input parameters comprise a mass flow of a gas from the heat source or a quantity that is representative for this mass flow and/or a temperature of the exhaust gas and/or a temperature of the outside air and/or a condensation temperature of the working medium of the cyclic process.

8. Method according to claim 1 wherein the control of the thermal cyclic process is carried out by means of model predictive control (MPC).

9. Method according to claim 1 wherein the control of the thermal cyclic process takes place by means of setting a rotational speed of a feeding pump and/or of an expansion machine and/or of a condenser fan.

10. Method according to claim 1 wherein the control comprises a repeated or continuous comparison of an actual value with the setpoint value of the process variable.

11. Device comprising: a thermal cyclic process, in particular an ORC, with one or more heat exchangers for transferring heat from one or more heat sources to a working medium of the thermal cyclic process; and a microprocessor for carrying out the method according to claim 1 for controlling the thermal cyclic process.

12. Computer program product comprising at least one computer-readable medium with instructions that can be executed by a computer and that are for carrying out the steps of the method according to claim 1 during operation on a computer.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0039] FIG. 1 shows an ORC system schematically.

DESCRIPTION OF THE EMBODIMENTS

[0040] FIG. 1 shows, in a schematic manner, a system which contains an Organic Rankine Cycle 100.

[0041] A pump/feeding pump 10 brings the fluid working material to operating pressure. When the fluid working medium flows through a vaporiser 20, heat from a heat source 30 is fed to the working medium. Due to the application of energy, the working medium is pre-heated and partially or fully vaporised and, where applicable, superheated. As a rule, saturated vapour or wet vapour or superheated vapour forms at the outlet of the vaporiser 20. The vapour of the working medium flows out of the vaporiser 20 via a pressure pipe to the expansion machine 40, where it is expanded to a lower pressure, as a result of which work is performed which is converted into mechanical energy by the expansion machine 40 and then, for example, by means of a generator coupled to it, further converted into electrical energy, or used in another form. e.g., by the direct drive of mechanical consumers or of the drive of a hydraulic pump. The vapour then flows through a condenser 50, where the vapour gives off the sensible heat and the condensation heat to a cooling cycle with a cooling medium flow. The working medium condenses and changes over completely into the liquid state of matter. The saturated fluid or supercooled working medium is stored temporarily in the storage container 60. The pump 10 then brings the saturated working medium from the storage container 60 back to operating pressure, and consequently closes the cycle.

Approach of the Invention

[0042] The object of the invention is a concept for controlling the Organic Rankine Cycle 100 or also another cyclic process in conjunction with a highly dynamic waste heat source 30. Very generally understood as a dynamic waste heat source is a heat source whose heat output fluctuates over time due to a change in the mass flow holding the heat and/or in the temperature and/or in the thermal capacity. For example, particularly called dynamic or highly dynamic are those heat sources in which a heat output H1 at a time t1 can change by more than 25%, more than 50% or even more than 75% of the heat output H1 to a heat output H2 at a time t2, whereby the time difference t2t1 can be less than one hour, less than ten minutes, less than one minute, less than one second or less than 0.1 second.

[0043] The ORC system consists of different components that have a different influence on the efficiency. The components consequently have a different, usually non-linear development of the efficiency in a case of a partial load, and several correcting values have contrary influences on the efficiency. Cited here as an example is the condenser 50. If the condenser 50 is better cooled due to a higher airflow rate, the efficiency of the cyclic process increases, i.e., it Is possible to convert more energy (P.sub.el, gross). In contrast, however, there is a rise in the auxiliary power (P.sub.el, auxiliary power), which includes the energy to be expended for the fan. A further extremely important component is the heat exchanger/vaporiser 20, in which the waste heat is transferred to the ORC working medium. Due to the controlling interventions in the feeding pump/pump 10 and the expansion machine 40, the heat output {dot over (Q)}.sub.TransferredHeat transferred to the ORC 100 from the waste heat source 30 by the heat exchanger 20 changes significantly.

[0044] The control concept according to the invention makes it possible to drive the ORC 100 in the optimal operating point, whereby the optimisation criterion is the so-called system efficiency .sub.System, which is made up of the efficiency of the heat transfer .sub.HT and the cycle efficiency .sub.th, net of the ORC 100. These two partial efficiencies are not independent of each other. Consequently an increase in the cycle efficiency can worsen the heat transfer efficiency to the point that the product of the two efficiencies

[00001]${}_{\mathrm{System}}={}_{\mathrm{HT}}.Math.{}_{\mathrm{th},\mathrm{net}}={}_{\mathrm{HT}}.Math.\frac{\text{?}-\text{?}}{\text{?}}$$\text{?}.Math.\text{indicates text missing or illegible when filed}.Math.$

is nevertheless reduced.

[0045] According to the invention, the control of a thermal cyclic process, in particular of an ORC with model-based control, is carried out in such a way that operation optimised with respect to the system efficiency occurs. One of the problem areas that is solved according to the invention is that an online optimisation calculation cannot be represented with sufficient speed because this calculation would have to take into account a multitude of parameters.

[0046] Available as input quantities are the mass flow of the exhaust gas from the heat source (or other quantities that allow the mass flow to be designated, such as, e.g., a rotational speed of a combustion engine) and its temperature, as well as the temperature of the outside air. The rotational speed of the feeding pump, the rotational speed of the fan motor of the condenser and the rotational speed of the expansion machine can be changed, whereby the effects of changes are complex to some extent (e.g., heat transfers, change in the pressure situation at which the vaporisation takes place, changeable areas for pre-heating and vaporisation) and can frequently be calculated only iteratively.

[0047] According to the invention, the controller should not carry out an optimisation for a given load point during ORC operation, and instead should already know, based on stored data, at which rotational speeds the individual assemblies should run in order to obtain the maximum energy yield with the current parameters of the heat source. The approach is now such that the system is mapped in a simulation environment whereby the actual behaviour of the components in the event of a load change is mapped. The behaviour of the components can be derived from trials or it can also be known from the literature (model equation, correlations).

[0048] For example, now the optimal vaporisation temperature can be calculated in dependence on the mass flow and temperature of the heat source as well as on the condensation temperature. Now a smoothing function (e.g., a polynomial) is defined for the optimal vaporisation temperatures that have been found, whereby this smoothing function makes it possible to determine the vaporisation temperature directly and without the model equations for the individual components from the parameters of the heat flow and the parameters of the condensation. The vaporisation temperature calculated in this way can now be used as a correcting value for the MPC.

[0049] In the first step, the control concept, working with a pre-determined combination or mathematical mapping between values of the input parameters and values of the process variables, consequently calculates which values of the process variable or of the process variables (setpoint values) allow optimised operation and then adjusts the rotational speeds of the feeding pump, expansion machine and condenser fan such that these process variables are achieved quickly and reliably. By means of a continuous comparison of the setpoint values and the actual state, the reaction here is also quick, whereby the future actual state is also taken into consideration in the control. The average relative system efficiency is up to 10% greater than in the case of conventional control with fixed vaporisation temperatures.

[0050] In summary, it can be recorded that while model predictive control with fixed setpoint values (e.g., fixed-pressure control) according to the state of the art allows fast control, no optimisation of the operating point takes place in this control. In addition, online optimisations are very computation-intensive and too slow for highly transient processes. In contrast, the invention's designation of the optimal process variables ahead of time and the determination of a function for calculating the process variables depending on the input parameters are fast and economical. A combination of these two concepts allows simple, economical control that is optimised with respect to the system efficiency and that allows, without additional costs, an energy yield that is approximately 10% greater than that of conventional concepts.