Method for operating at least one pump assembly of a multitude of pump assemblies

Abstract

The method serves for operating at least one pump assembly (1, 1a, 5) of a multitude of pump assemblies (1, 1a, 5) which each comprise a programmable electronic motor (4, 4a, 6) control for individual pumps or groups or pumps, and which are at least temporarily data-connected to a server (8) via a network, with which data, in particular parameters and/or operating data of the multitude of pump assemblies (1, 1a, 5) is transferred via the network to a server (8) and is stored there in a data base. The stored data is processed according to defined typically statistical computation rules, whereupon the at least one pump assembly is then adapted in its programming via the network, on the basis of the processed data.

Claims

1. A method for operating at least one pump assembly of a multitude of pump assemblies, the method comprising the steps of: providing the pump assemblies so as to each comprise a programmable electronic motor control for individual pumps or groups of pumps and which are at least temporarily data-connected to a server via a network; transferring data comprising parameters and/or operating data of each of the multitude of pump assemblies via the network to the server and storing the transferred parameters and/or operating data in a data base associated with or comprised by the server; processing the stored parameters and/or operating data according to defined computation rules; and adapting programming of the at least one pump assembly, via the network, based on the processed data, wherein, with a first group of pump assemblies, operating data is determined by way of electrical variables of a drive as well as by way of at least one detected hydraulic variable, and with which with a second group of pump assemblies of the same type, operating data is acquired by way of electrical variables of a drive, and on the server side, the operating data of the second group of pump assemblies is adapted in the programming on the basis of the detected hydraulic variables of the first group.

2. A method according to claim 1, wherein: the multitude of pump assemblies include assemblies of different types; and the step of transferring parameters and/or operating data includes acquiring data of the pump assemblies of different types in the data base and further comprising grouping data of the pump assemblies of different types for processing.

3. A method according to claim 2, wherein the operating data of pump assemblies of the same type is statistically evaluated, and the programming of at least one pump assembly is adapted based on a predefined deviation from the evaluated data.

4. A method according to claim 1, wherein the pump parameters are provided as pump curves of a multitude of pump assemblies or pump assembly groups and are statistically evaluated on the basis of the operating data, stored in the data base, of pump assemblies of the same type or of pump assembly groups of the same type, wherein the results of the statistical evaluation comprise mean data which is used for programming a pump assembly or a pump assembly group.

5. A method according to claim 1, wherein data of an operating condition of a pump assembly or of a pump assembly group is compared to data from the data base, of operating conditions which are the same with regard to the hydraulic situations, of pump assemblies or pump assembly groups, of the same type, wherein an in particular energetically more favorable operating condition is determined by the comparison, and the pump assembly or the pump assembly group is programmed for reaching the more favorable operating condition.

6. A method according to claim 5, with which, with a pump assembly group, switching points for connecting and disconnecting the pump assemblies of the group are programmed.

7. A method for programming a programmable, electromotorically driven pump assembly or a pump assembly group, the method comprising the steps of: providing pump assemblies or pump assembly groups or both pump assemblies and pump assembly groups that are programmable electromotorically driven pump assemblies of the same type and in operation; transferring parameters and/or operating data of each of the pump assemblies or each of the pump assembly groups or both each of the pump assemblies and each of the pump assembly groups to a cloud-based data base; comparing parameters and/or operating data of the pump assembly or of the pump assembly group with data stored in the data base; and adapting programming of the pump assembly or of the pump assembly group based on a predefined deviation determined during the comparing step, wherein, with a first group of pump assemblies, the operating data is determined by way of electrical variables of a drive as well as by way of at least one detected hydraulic variable, and with which with a second group of pump assemblies of the same type, the operating data is acquired by way of electrical variables of a drive, and the operating data of the second group of pump assemblies is adapted in the programming on the basis of the detected hydraulic variables of the first group.

8. A fluid pump system comprising: a multitude of pump assemblies having a programmable electronic motor control for individual or groups of the pump assemblies; a server with a data base for storing parameters or operating data or both parameters and operating data of each of the pump assemblies; a network; a data connection at least temporarily data-connecting to the server via the network; and a data processing device configured to evaluate data transferred from each of the pump assemblies, according to predefined rules, and to adapt programming of at least an individual pump assembly based on the evaluated data, wherein: the multitude of pump assemblies comprises a first group of pump assemblies which is provided in each case with at least one sensor for detecting a hydraulic variable and the operating data is determined by way of electrical variables of the drive as well as by way of at least one detected hydraulic variable; the multitude, of pump assemblies further comprises a second group of pump assemblies of the same type and the operating data is determined by way of electrical variables of the drive; and at the server, the operating data of the second group of pump assemblies is adapted in the programming on the basis of the detected hydraulic data of the first group of pump assemblies.

9. A fluid pump system according to claim 8, wherein the evaluation is effected according to statistical rules.

10. A fluid pump system according to claim 9, wherein the network is internet based and the pump assemblies or pump assembly groups comprise a webserver or are connected to the internet via a LAN or WLAN.

11. A fluid pump system according to claim 8, wherein the network is internet based and the pump assemblies or pump assembly groups comprise a webserver or are connected to the internet via a LAN or WLAN.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) In the drawings:

(2) FIG. 1 is a greatly simplified representation, the linking of a multitude of pump assemblies, by way of a data base;

(3) FIG. 2 is a flow diagram, from the detection of pump data up to the reprogramming of a pump assembly;

(4) FIG. 3 is a diagram showing the power consumption of a pump assembly consisting of several pumps, at a different transition speed; and

(5) FIG. 4 is a diagram showing three pump curves of pump assemblies of the same type.

DESCRIPTION OF THE PREFERRED EMBODIMENTS

(6) Referring to the drawings, the fluid pump system which is symbolically represented in FIG. 1 comprises a pump assembly 1 with an electric motor 2 driving a centrifugal pump 3 and comprising an electronic motor control 4 which is part of an electronic frequency converter activating the electric motor 3. The electronic motor control 4 is programmable.

(7) The fluid pump system which is represented in FIG. 1 comprises three pump assembly groups 5. Each pump assembly group 5 consist of three pump assemblies 1a, each driven by an electric motor 2a driving the centrifugal pump 3a. There, an electric motor control 4a which is programmable is also assigned to the electric motors 2a in each case, to which motor control a central electronic control 6 which is likewise programmable is superordinate. The centrifugal pumps 3a are hydraulically connected in parallel and the pump assemblies 1a are connected or disconnected by the electronic control 6 depending on the power demand, so that the pump assembly group 5 in the system practically forms one pump assembly, as is usual with such booster pumps.

(8) The pump assemblies 1 and 1a as well as the pump assemblies groups 5 and which are represented in FIG. 1 are specified here only by way of example, and in practise such a fluid pump system consists of thousands of pump assemblies and pump assembly groups, said assemblies and groups being of the same type, but can also be of a different type.

(9) The pump assembly 1 as well as the pump assembly groups 5 are wirelessly connected to a network, as indicated by the data connection arrows 7. The connection to a cloud-based data base is effected by way of the internet, by which means pumps throughout the world have access to this data base. The data base forms part of a server 8 which can either be software-implemented or is formed by one or more computers with a suitable read-only memory, which are specifically provided for this purpose and on which a data base program runs. In this context, an existing data base of the applicant, the Grundfos Hydrobooster MPC is referred to, wherein this acquires the data of the pump assemblies which are connected thereto and its data can be retrieved by an operating person 9, for example by way of a smartphone 10, on which a suitable software application runs. The smartphone 10 is also provided for direct communication with the electronic control 6, which is to say is provided for and is suitable, for programming and for setting the control.

(10) The cloud-based data base, thus the server 8, can be realised for example by a platform which is known under the trademark Azure, from the company Microsoft and which is capable of receiving data from different pump assemblies 1, la and pump assembly groups 5, processing this data and as the case may be, after the processing, of sending it to a pump assembly 1, la for the purpose of programming.

(11) The data which is transferred to the server 8 can be current operating data, such as motor current, motor voltage, speed, measured flow rate, calculated flow rate, measured differential pressure, calculated differential pressure, operating time, etc., but also data which does not change, such as for example the serial number of the pump assembly, the number and serial number of the sensors, the length and diameter of the connected conduits, locations of the facility and the like. The data base of the server 8 thus comprises a multitude of different data of different pump assemblies 1, la and pump assembly groups 5.

(12) On the server side, the processing of the data is effected in a first step 15, in which the data sets of the individual pump assemblies are received. In a step 16, the data which is received from the different pump assemblies and pump assembly groups is assigned, and specifically in groups according to pump types. Thus for example there is a pump type which comprises booster pumps, which is to say pump assembly groups 5 consisting of several individual assemblies 1a which however are activated and programmed by a common electronic control 6. The multi-stage pumps for example form a further group, and another group in turn is assigned to the heating circulation pumps. Thereby, the groups do not necessarily need to be grouped according to the aforementioned manner, and such a grouping is effected as would appear most useful with regard to the evaluation. Thus for example the pump assembly groups 5 can be assigned to one data group, whereas their pump assemblies 1a can be assigned to another data group. Thereby, it is useful to bring together that data of the pump assemblies which is also to be used at a later stage for a statistical evaluation or for a comparison.

(13) The respective data sets are then examined with regard to consistency in step 17, wherein a further examination as to whether the transferred operating data lies within a previously determined average tolerance band is effected. Data sets which exceed this tolerance band are discarded. In step 17, the data which is assigned to a group of pump assemblies is normalised, which is so say further processed such that pump assemblies of the same type but of a different construction size can be compared to one another.

(14) The data transferred from the pump assemblies, in step 18 is then processed with the help of statistical methods and/or mathematical pump models, and the data which has been revised with regard to the data base inasmuch as this is concerned, in step 19 is transferred to a pump assembly 1, 1a or to a pump assembly group 5, at which the existing data is replaced by this processed data.

(15) With the assignment of the data sets to pump groups, these data sets do not necessarily need to be assigned to only one group, but the formation of sub-groups is also conceivable. It is also possible to form the pump groups such that multiple overlapping are envisaged. Thus for example pump assemblies 1, la or pump assembly groups 5 which for example comprise a wearing part of the same type, for example a shaft seal, can be assigned to one group. On the other hand their assignment according to the delivered medium is also conceivable.

(16) As to how such a grouping can look is explained hereinafter:

(17) Table A includes multi-stage pump assemblies of the CR construction series whose serial number is acquired, whose field of application is acquired and with which the time between consecutive maintenance intervals is acquired. Moreover, with part of these pump assemblies, the delivery rate is detected by sensor, and this is stored in the table as data pairs (point in time, flow rate) of the point in time of the measurement and the magnitude of the flow rate.

(18) TABLE-US-00001 TABLE A time between measured consecutive delivery rate serial name of the field of maintenance (point in time number pump application intervals (hours) l/hour) 18495 CR6 industry 4310 10:00, 1990 11:00, 1895 12:00, 1995 12112 CR6 industry 2950 13:00, 2210 14:00, 2190 15:00, 2220 13180 CR6 industry 3512 no data 7514 CR6 home 1620 08:15, 1900 08:30, 1910 08:45, 1902 10712 CR6 industry 4140 15:00, 1850 16:00, 2209 17:00, 2300 8212 CR6 home 1770 no data

(19) As the following table B shows, multi-stage centrifugal pump assemblies of the CR construction series are also specified there, with which the serial number is specified, as well as the type of the shaft seal, the expected service life of the seal and the delivered medium.

(20) TABLE-US-00002 TABLE B expected service serial name of the life of the seal delivered number pump type of shaft seal (hours) medium 1255 CR5 Typ P 12050 water 1421 CR5 Typ P 1800 water 17643 CR6 Typ O 600 glycol 13210 CR4 Typ O 800 glycol 8212 CR5 Typ P 9100 water 1975 CR4 Typ O 2100 oil

(21) As the two tables A and B which each represent a group illustrate, a sorting according to different criteria can be effected.

(22) Finally, the following Table C shows heating circulation pump assemblies of the type series Alpha 1, Alpha 2 and Alpha 3, each with their serial number, the type of their regulation, the current differential pressure at a given point in time, and the measured delivery rate at a given point in time.

(23) TABLE-US-00003 TABLE C differential measured name pressure delivery rate serial of the type of (point in (point in time, number pump use regulation time, m) l/hour) 1534 Alpha 1 heating proportional 12:00, 3.2 12:00, 300 12:30, 3.6 12:30, 305 13:00, 4.0 13:00, 320 8422 Alpha 2 heating proportional 12:00, 4.1 12:00, 360 12:30, 4.1 12:30, 360 13:00, 4.1 13:00, 360 21987 Alpha 1 heating constant 04:27, 2.9 04:27, 200 pressure 04:28, 2.8 04:28, 201 04:29, 2.9 04:29, 199 77865 Alpha 1 heating proportional 06:00, 3.3 06:00, 267 12:00, 3.3 12:00, 267 18:00, 3.5 18:00, 271 5423 Alpha 3 heating constant 22:10, 3.6 22:10, 295 pressure 22:30, 3.6 22:30, 291 22:55, 3.7 22:55, 298 53142 Alpha 1 heating constant 08:00, 4.8 08:00, 1477 pressure 09:00, 4.7 09:00, 1600 10:00, 4.7 10:00, 1755

(24) The grouping is effected in method step 16.

(25) The principle of the method according to the invention is thus based on evaluating the data of a multitude of pump assemblies or pump assemblies groups, which for example are already in operation throughout the world and whose data is acquired in a cloud-based manner, such that the operation of individual pump assemblies or pump assembly groups can be improved by way of this, or, with renewed starting operation, a programming of the electronic motor control can be effected on the basis of comparable pump assemblies/pump assembly groups which are already running and are in operation. Thereby, the pump curves of a multitude of pump assemblies of the same type or pump assembly groups of the same type can be statistically evaluated, for example by way of evaluating the server-side data, in order to bring these pump assemblies and/or pump assembly groups as close as possible to the actual operating points. As is shown in FIG. 4, pump curves a, b and c result for three pumps of the same type, and these curves are distanced to one another and do not correspond to one another. A pump curve which results from the mean and which has been determined on the basis of a multitude of pump assemblies is stored in the programming of an individual pump assembly after the statistic evaluation of a multitude of such pump curves.

(26) As has been described above, with pump assemblies of the same type, of which some are operated with a hydraulic sensor and some without a hydraulic sensor (pressure sensor or flow sensor), not only can the method according to the invention be used to render the model computation significantly more precise by way of comparing the operating points computed by model on the basis of the electrical values, with the sensorically determined operating points, but the pump curves forming the basis of the pump assemblies can be put on a significantly broader statistical footing than would be possible with the more complicated, but in comparison fewer laboratory measurements.

(27) The flow rate through the pump, thus the temporal course of the delivery volume flow, in the case of pump assemblies without a sensor, is determined on the basis of the following equations:
p=a.sub.h2q.sup.2+a.sub.h1qn+a.sub.h0n.sup.2(1)
P=a.sub.t2q.sup.2n+a.sub.t1qn.sup.2+q.sub.t0n.sup.3+B.sub.1n.sup.2+B.sub.0n+P.sub.0(2)
with

(28) TABLE-US-00004 p differential pressure between the pump inlet and the pump outlet or the pump group inlet and pump group outlet (booster pumps) P electrical power q flow rate through the pump n speed of the pump a.sub.h2, a.sub.h1, a.sub.h0 parameters of the mathematical model representing the pump pressure a.sub.t2, a.sub.t1, a.sub.t0, B.sub.1, B.sub.0, P.sub.0 parameters of the mathematical model representing the pump power

(29) The flow rate q can be computed from the equation (1) on account of the pressure difference p of the pump or of the pump group and the speed n. It is possible to compute the flow rate q from the equation (2), from the pump power P and the pump speed n. These computations depend on the parameters a.sub.h2 to a.sub.h0 and a.sub.t2 to a.sub.t0. If the power P as well as the differential pressure p and the speed n are known, the flow rate q can be computed as follows

(30) $\begin{array}{cc}q={}_1\frac{P}{n^2}+{}_2\frac{p}{n}+{}_{3}n+{}_4+{}_5\frac{1}{n}+{}_6\frac{1}{n^2}& \left(3\right)\end{array}$

(31) TABLE-US-00005 .sub.1, .sub.2, .sub.3, .sub.4, .sub.5, .sub.6 parameters of the model representing the flow rate, whilst using the aforementioned differential pressure models and power models

(32) The flow computation depends on the parameters .sub.1 to .sub.6, which are typically determined on the part of the manufacturer by way of tests. This work is time-consuming and expensive. For this reason, these parameters are mostly only available for a few pumps and are interpolated for others. As is shown in FIG. 4, which represents three pump curves a, b and c of three different pumps of the same type, these curves have a scatter e between the pump a and the pump b. Considering now pumps a and c, this scatter is even larger. If therefore only one pump, for example pump b is measured, then a significant scatter in the pump curves results even with pumps of the same type. In order to reduce such scatter, according to the invention, one envisages determining the parameters of the above equations on the basis of large data quantities as are available on the data base side at the server 8, of a large number of pump assemblies, on account of the multitude of pump assembles of the same type and by way of statistical evaluation, in a comparatively precise manner and specifically where possible, whilst using the data of pump assemblies, with which the differential pressure p and/or the volume flow q are determined by sensor. The equation (2) can be improved with this information, since the power and the speed of the pump are always available on account of the corresponding motor data, so that parameters which describe the average electrical/hydraulic behaviour of the pump assembly type concerned can be found. The parameters of the equations (1) and (3) can be determined if a flow sensor is also present additionally to the differential pressure sensor. The determining is effected by way of known mathematical methods, for example by way of linear regression. Thereby, a continuous updating of the parameters of the equation (1) is effected. The sought parameters are collected as parameter vectors and are updated with known methods, so that it is not necessary to compute these afresh each time from the complete data, when a parameter set is requested.

(33) Since it is particularly with the wireless communication that a loss of data packets or files can occur, at all events one must examine whether the applied data set is consistent before a parameter updating. Thereby, the data set to be used is applied in the already available model and it is examined as to whether the value determined with this lies within the standard deviation of the applied model. The data set is not used for further determining the parameter if this is not the case. Thereby, it makes sense to not only use pump assemblies of the same type for determining the parameters, but where possible also those which are used in a comparable manner and which roughly have the same operating hours, in order to be able to compensate any occurrence of wearing.

(34) The method according to the invention can also be used for example to determine the energetically most favourable switching points in the case of booster pumps with which a group 5 of pump assemblies 1a are operated in parallel and with which the pumps are connected and disconnected according to requirements, by way of the common electronic control 6. The connecting is thereby often effected according to simple rules, for example the next pump is connected when the already running pumps have reached 95% of their maximal speed. This is often not energetically favourable.

(35) According to the invention, the data for example of three pump assembly groups is evaluated with regard to their temporal course, as is represented by way of FIG. 3. The three pump assembly groups have different switching speeds the pump assembly group 1 (pump station 1) has a switching speed n.sub.1 which is lower than the switching speed n.sub.2 of the pump assembly group 2 (pump station 2) which in turn is lower than the switching speed n.sub.3 of the pump assembly group 3 (pump station 3). If one now compares the power consumption P over time T, in dependence on the running pump speed s, then it becomes clear that the pump station 2 has the energetically most favourable switch-over speed n. This switching threshold which is effected by way of evaluating operating data on account of the data base 8 can be effected in an automated manner and serve for adapting the switch-over speed of such pump stations, be it on installation or after a certain time. The switching speed n must always be determined afresh if the inlet pressure significantly changes, since these settings are dependent on the pressure conditions at the inlet of the booster pump. One then seeks comparable data of booster pumps in the data base, and the most favourable switch-over speed for the pump station is computed afresh on the basis of this comparison which is effected similarly as described by way of FIG. 3.

(36) While specific embodiments of the invention have been shown and described in detail to illustrate the application of the principles of the invention, it will be understood that the invention may be embodied otherwise without departing from such principles.

APPENDIX

List of Reference Symbols

(37) 1, 1a pump assembly 2, 2a electric motor 3, 3a centrifugal pump 4, 4a electronic motor control 5 pump assembly group 6 electronic control of 5 7 data connection path 8 server 9 operating person 10 smart phone 15 to 19 method steps s pump speed n switching speed e deviation (FIG. 4)