Hydraulic Actuator for Valves in Brewery, Dairy and Pharmaceutical Industries

Final Report Summary - HYDRACT (Hydraulic Actuator for Valves in Brewery, Dairy and Pharmaceutical Industries)

Executive Summary:
The aim of the project 'Low-Energy Hydraulic Actuator for Valves in the Brewing, Dairy and Pharmaceutical Industries' (HYDRACT) is to develop motion technologies based on water hydraulics. The target is to replace the pneumatic technology which has been the dominating technology for actuators for the last 4 to 5 decades with water hydraulics.

The obvious and short term motivation for the sanitary industry to embrace the water hydraulic technology is to reduce energy costs and CO2 emissions for operating actuators by up to 70%. This is possible because of the higher system efficiency of water hydraulics and by avoiding leaking pneumatic systems. The hydraulic media which is sterilised tap water mixed with glycol is incompressible in the operative pressure range between 50 bar and 100 bar. This gives the system a reliability and stability allowing the industry to trust the valve to stay in the requested position - normally open or closed. A direct benefit for the industry is that unintended mix of products following a pressure peak (water hammer) will not occur. The reduced energy consumption and the reduced waste represent a significant reduction in the running costs for the industry.

The higher reliability, precision and power density over pneumatic actuators will also enable the industry to design future installations differently by using sanitary valves as regulation valves with unheard precision. The valve can be trusted always to stay in the wanted position because of the incompressibility of water. The perspective is therefore to change from matrix grid systems towards bus systems. This will simplify the systems and reduce the capital investment related to pipe, valves and pumps by 25% to 40%.

The consortium has developed the hydraulic system to perform all the known functions of any type of sanitary valve including regulation valves. Wireless communication technology and relevant sealing technology suitable for water has been developed, tested and verified in field tests at Carlsberg in Fredericia who is participating in the project as end user.

The first step in the plan for exploiting Hydract' business potential is to launch the hydraulic actuator for retrofitting onto existing valves. This allows the customer to re-use the existing control tops from the valves and thereby reducing the investment and avoiding reprogramming of the customers mainframe control system. This is expected to facilitate building up the necessary level of confidence at the customer. The next step is to launch complete actuators including control units applicable to any of the normal communication protocols used in the industry - both wired and wireless.

The consortiums approach regarding dissemination changed from being very open towards being very restrictive for competitive reasons. The supply chain is focussed on KM Rustfri who holds the IPR for the patented invention. The other SME's Norsistemas (Portugal) and GFD (Germany) will act as suppliers to KM.

The consortium will launch the Hydract actuator for retrofitting on the Brau Beviale tradeshow in Nuremberg in November 2011 supported by Carlsberg. The launch into the dairy and pharmaceutical industries will take place at Anuga Food Tech in February 2012 and Achema in June 2012.

The consortium has furthermore applied for additional financial support under the FP 7 program for demonstration actions to fund additional field tests and thereby supporting a successful market introduction for complete actuators.

Project Context and Objectives:
The first phase of the project defined the functional envelope of the complete actuator system including the hydraulic system, seals, control and communication network. Information came from benchmarking data of existing equipment, end-user discussions, partners' experience, and standards.
Four design and development packages were then run in parallel to produce prototype equipment for end-user trials towards the end of the project. BHR were responsible for the hydraulic system design and actuator development, while LUH and UM were responsible for seal development and communication/control respectively.
The hydraulic system was designed to be flexible and could operate between 5 to 50 actuators depending on the duty cycle. The system was modelled using a computer simulation package and guidelines were written for operation and maintenance in the sanitary valve application within the Carlsberg Brewery at Frederica, Denmark.
The actuator was initially designed by BHR, and revised by KM/LUH to account for manufacturing and seal issues. LUH purpose built a seal test rig within the project, but the production GFD supplied PTFE seals were used in the early prototype actuators.
UM and NS designed and supplied the radio frequency communication system that operated between the actuator hydraulic power unit. They were also responsible for the development of an internal magnetic position sensor that proved to be compromised by material choices for the actuator cylinder.
BHR assembled and tested the prototype system at their laboratories using a purpose built performance/endurance rig. The first actuator design met the majority of design targets but was overly complex to assemble and prone to internal seal leakage.
At month 19 the end-user requested design changes to the actuator for the site demonstration that required a remanufacture of the prototype unit and changes to the hydraulic system. An initial site demonstration was conducted with a simplified actuator that could be retro-fitted to an existing valve and communications module. This trial was very successful, enabling the end-user to appreciate the benefits of the water hydraulic option firs hand. The equipment operated with no problems for a two-week period and demonstrated an energy saving of 65% when compared with the installed pneumatically operated system.
A 6-month project extension was granted to prepare for and conduct a second demonstration at the Carlsberg site using the requested multi-function actuator with a regulating capability. The actuator was detailed and manufactured by KM, while BHR designed the hydraulic control system and modified the hydraulic power unit. The actuator was assembled and tested at BHR with the new hydraulic control system before being shipped to Denmark for the second trials.
The December 2011 trials proved the double seat lift functionality of the actuator and also convinced Carlsberg engineers that the valve could be used as a regulating valve within a closed-loop control system.
The HYDRACT FP7 project and actuator system have been publicised at the IFPEX 2010 exhibition in Birmingham UK, and in the BFPA (British Fluid Power Association) 2010/11 Yearbook

In this summary we have decided to highlight some of the objective and results. One big task was to make a benchmarking report on defining product features required by industry including expected operating life time and servicing intervals. Including Specification of size, range and type of valves, target manufacturing cost/sales price, servicing costs/failure rates, operating speeds (time) and number of cycles. This report funded the basic for the future work in the other WPs. In continuation of the report the task of identifying electrical power and signal transmission requirements and standards for hydraulic system components including solenoids, valve sensors, instrumentation and RFBT components was worked on in the first reporting period. For the prototype also studies based on pressure data and force data were performed.
The end-user has been involved in the process to ensure that their requirements also were meet in the development.

The final product will have clear advantages to current pneumatic process machinery in these industries, especially having in mind, performance, food safety and energy savings.

At the end of the project a prototype will be developed which can be fitted on to the majority of valves used in the dairy, brewery and pharmaceutical sector. This will demand that connecting devices coupling valve and actuator are developed wherefore it may be an idea to target for the one or two leading valve brands implemented in industry.

The project builds on ideas for which patents applications have been filed. Before the end of the project the PCT phase will be completed and the national/regional applications will be issued. This ensures a strong business platform after market introduction of the product and the connecting systems.

Project Results:

1.3.1 Functional Envelope Definition and Calculation

1.3.1.1 Actuator Specification

Specifying the actuator requirements was viewed as the most important part of the initial work package, and drew upon existing valve standards as well as end-user requirements from Carlsberg, Fredericia. A specification evolved around mix-proof and leak-proof designs used predominantly in liquid food applications, and a decision was made to focus on the 3" (DN80) valve as this was the most commonly used size.

1.3.1.2 Specification of Fluid

The initial intention was to use sterile water as the working fluid for the hydraulic system, although in reality the choice was far from simple and based upon tribology, hydromechanics, environment, hygiene and maintenance. Sterile water is in fact expensive to produce and maintain and can have a detrimental effect on system components. Alternatives such as tap water proved to be much more practical with a fairly good track record for reliability over the past decade.
When problems are likely to be encountered with freezing temperatures or poor lubrication, then water-glycol is acceptable and commonly used in breweries/process plant for heat exchange. Food grade oils are another option and have been approved for indirect contact with foodstuffs throughout the world.
To aid in the final choice of fluid a weighting analysis was completed by round-table discussion at BHR. This confirmed that tap water and a high concentration water-glycol were the best fluids for use in the HYDRACT system.

1.3.1.3 Seal Type Assessment

The material screening for seals for use with water based hydraulic fluids showed different possible material solutions. First investigations had shown that water as a hydraulic fluid has special requirements on the seal material. From the chemical point of view it is very reactive and can cause hydrolysis reactions and soaking of the seal material. Especially the use of polyurethanes, polyacrylates and NBR-elastomers materials could be problematic. The poor lubricity of water based hydraulic systems was another issue that needed to be considered. Stiction associated with rubber materials is a common problem and would require laboratory testing if chosen for the seal material.
PTFE, PEEK and PE looked more promising with good chemical resistance against water based media and suitable tribology qualities.
PTFE, PEEK and PE seals can also be produced in a turning process, which allows a fast availability of prototypes and gives the opportunity for a quick variation of sealing geometries.

1.3.1.4 Hydraulic Circuit

Initial calculations showed that it would technically feasible to provide a hydraulic solution for both industrial valves and the lower duty process valves. However, the key to success for HYDRACT would be to provide a smart design that was energy efficient compared to existing installations.
A concept was proposed that utilised a centralised hydraulic supply unit, manifold and valves supplying remotely located actuators. Pressure drops seemed reasonable with relatively small bore pipework, which should lead to an unobtrusive installation. Using a pump that trips in and out based on the system demand seemed an obvious choice to save energy.
A detailed knowledge of valve operating sequences/duty cycle would be used to optimize the manifold and valving design, hence ensuring efficiency in operation.
The added complication of flushing mix-proof or leak-proof valves was well within the capabilities of the hydraulic option, and initially a single seat lift function was specified by Carlsberg.
The D1.2 report discussed cleanliness, safety and reliability issues, all of which appeared to be acceptable for the application, certainly in brewery and dairy industries.

1.3.1.5 Control System Assessment

An assessment was made of the main criteria for a valve position sensor and the various commercially available options. UM proposed the best solution would an electromagnetic sensor mounted on the side of the actuator cylinder, with a corresponding magnet within the piston. In this solution there are no additional penetrations required into the high pressure envelope of the actuator.
A control strategy was needed between the valve and main system whereby not only valve position, but solenoid status and switching are inputs communicated. Any discrepancies in operation would need to be identified to flag up faults and maintenance issues.
Communication formed a major part of the control system and was based on RF technology to save cost and simplify installation. Battery powered devices were proposed and hence needed to have a low power consumption to maximize battery life.
Of the various RF protocols available, ZigBee had the following advantages:

● 1mW power consumption
● Indoor range 30m
● ZigBee-Pro 60m range indoor/ 750m outdoor
● Relatively cheap @ 15$/unit
● Immune to interference

UM conducted laboratory experiments with cell phone interference and metal shields around ZigBee transponders with no reported transmission failures.
Retrofit (or legacy) applications would require a wired protocol to the PLC such as:

● AS/I;
● Digital Interface;
● Device Net;

It was also suggested that the communication system must have an Application Programming Interface to allow system integrators to use the system.

1.3.1.6 Testing Programme and Test Rig Performance Specifications

Existing standards were firstly reviewed relating to the following areas:

● Fluid Power systems
● Hoses & pipework
● Cleanliness
● Valves
● Seals

A qualification procedure for the actuator has been suggested based upon tests that are generally applicable to pressure containing envelopes:

● Proof pressure
● Volumetric expansion
● Minimum burst pressure
● Impulse fatigue

Performance tests were also required to verify the pressure/force relationship and flow/speed relationship, while endurance tests would aim to prove long term reliability.

LHU proposed a seal test rig with synchron-servo-motor for precise motion control and force measurement using piezo transducers.

1.3.2 Systems Design

1.3.2.1 Notional System Design and Calculation Methodologies

The hydraulic system design ran in parallel with the actuator design as there were interdependencies between the two tasks. The initial actuator design required two pressure levels to operate the main stroke and the flush stroke of the actuator. The prototype hydraulic power unit was designed for a ten-valve system, with enough capacity to close five actuators simultaneously in the event of an emergency. The speed of operation of the actuators was required to be slow to prevent water hammer, and 30 seconds was suggested as the outside limit.
The hydraulic system design also had to take into account commercially available equipment that could meet the specification for the sanitary valve application. This narrowed the choice of suppliers to two in Europe, one in Denmark and one in the UK, and the project has used components from both suppliers during the course of the project.
The availability of a mini water hydraulic power pack capable of 140bar continuous delivery pressure effectively set the maximum operating pressure of the system. BHR proposed operation at this pressure to give the highest power density and compactness of the actuator, a decision that was subsequently revised in later versions of the actuator, where size was not deemed so important.

For the majority of the time the hydraulic circuit was required to provide a 100 bar line pressure to the distributed actuators, locking them in the closed or open position. The hydraulic pump was run up to the maximum operating pressure of 140bar, and hence a pressure reducing valve was required to drop the line pressure to the valves during normal operation.
The higher operating pressure of 140bar was required to flush the valve in the closed position, and this was applied to the actuators by by-passing the pressure reducing valve, allowing full system pressure into the distribution manifold.
A bladder type accumulator was located within the high pressure circuit and was pressurised by the pump up to the maximum operating pressure to maximise energy storage within the system. This was important in the event of a power supply failure or other peak flow demand.
Flow control valves were not used in the initial design as it was thought that the pressure reducing valve would provide the necessary flow restriction. In practice the actuator benefited from the slower operation afforded by additional flow control valves.
A pressure switch was used to turn the pump off and on in response to a flow demand, and this required a solenoid operated unloader valve to vent the pump pressure on start up.
The deliverable report for the hydraulic system design detailed the necessary calculations and components for the prototype circuit.

1.3.2.2 Design Methodology

The design methodology used for the HYDRACT system was formalised in a deliverable report that could be used by the SMEs for future installations. The process was simplified with the following steps;

● Define functional requirements/specification
● Identify/design hydraulic system components/function
● Draw initial system network
●Size main hydraulic system components based upon maximum pressure and flow requirements
● Re-evaluate hydraulic design based upon duty cycle
● Modify system network based upon duty cycle considerations
● Size hydraulic components based upon revised pressure and flow requirements
● Size hydraulic lines based upon pressure drop considerations
● Identify/design control system components/function
● Draw communication and control network
● Identify/design electrical power system components/function
● Draw power electrical distribution network

Sample calculations and guidelines were written for each step in the design process to make a comprehensive document.

A computer simulation was also performed on a network model of the HYDRACT system using the
Flowmaster simulation software. This enabled various operating scenarios and system configurations to be modelled on the computer in both steady state and transient modes.
The software tool provided a useful aid to training and would be valuable when adapting the HYDRACT system to different applications. The computer runs confirmed the early design calculations and clearly demonstrated the practicality of the system in multi-actuator and long line applications.

1.3.2.2 Maintenance and Installation Practice

The water hydraulic system had to be sufficiently clean to meet both the contamination requirements of the system components and the sterility requirements of the fluid for the given application.

Although 10 micrometer filtered tap water would meet the component cleanliness requirement, the brewery, dairy and pharmaceutical industry applications called for the working fluid that was 'sterile'. In order to achieve this, BHR Group limited investigated and devised its own procedures to ensure that a sterile system could be achieved and maintained in practice. This involved the addition of 0.45µm filters to remove organic contamination when filing the system and at the reservoir breather.

A detailed study was done on the type of contamination to be expected in the system, means of detecting microbial growth and a procedure to follow when testing hydraulic actuator hygiene. Dip slides were found to be a convenient means of detecting moulds and yeasts in the system fluid, and were used both in laboratory trials at BHR and during the site trials at Carlsberg.

Tests were performed on the prototype HYDRACT system at BHR so that a successful start-up procedure could be defined and verified for sterile applications. The following procedures were written;

● Procedure for filling system with hydraulic fluid
● Procedure for monitoring water contamination
● Prototype Hydract start-up procedure
● Cleaning procedure
● Flushing procedure
● Sterilisation procedure
● Actuator cleaning and sterilisation procedure
● System cleanliness checks during operation
● Accumulator maintenance

1.3.3 Actuator Development

1.3.3.1 Conceptual Designs

Conceptual designs for the hydraulic valve actuator were presented at meetings in Fredericia and Skive, in November 2009, and were subsequently superseded by a fully hydraulic concept.

Three concepts were produced based around a retrofit to the GEA Tuchenagen DN 80 mix proof valve:

● Concept # 1 - low risk extension of existing design
● Concept # 2 - smart hydraulic design but still with flush return spring
● Concept # 3 - fully hydraulic design

A weighted analysis was performed on the concepts based upon manufacturability, manufacturing costs, operational efficiency, operating range, number of components, and improved functionality.
The fully hydraulic concept was chosen primarily because it removed the troublesome return spring and was a design unique to the project.
It was important to consider the deep-drawing manufacturing capabilities of KM, especially when it came to designing the main body of the actuator. A finite element stress analysis was performed on the main body to identify stress raisers around the end cap and pressure ports, but eventually the design was driven by material thickness and L/D ratio constraints.

A conceptual design for a ¼-turn actuator was proposed based upon a rack and pinion adaption of the linear actuator, as this appeared to be the simplest approach. This was not followed through to detailed design or manufacture as the resources were deemed better directed towards the linear actuator development.

1.3.3.2 First Detailed Hydraulic & Engineering Design

BHR produced the first engineering design of the linear actuator in December 2008, detailing all the components for manufacture. The design was reviewed by LUH and GFD who outlined a major re-design to accommodate the uni-directional PEEK seals that they were proposing to use. The rigidity of the seals meant that key components had to be screwed together, and a double seal arrangement was necessary to give bi-directionality of sliding movement.
KM produced new engineering drawings based upon the revised sealing requirements and these were used for component manufacture of the first prototype actuator. Because of the tooling requirements for deep drawn components, it was decided to simplify the prototype by using commercially available tubing.

Provision was made for a position sensor magnet within the main piston, although cylinder material choices hindered the performance of the unit in practice.

1.3.3.3 Simplified Actuator Design

The first design of actuator underwent performance and endurance trials at BHR in preparation for site trials at the Carlsberg Brewery, Fredericia.
However, during a meeting with Carlsberg prior to the trials it became obvious that they wanted the actuator to be retro-fitted to a Südmo DN80 valve, with communication to the plant through the valves 'Intellitop' control unit. This valve was of the double-seat lift design and hence also required an additional lower seat lift function from the actuator, something that required an extensive re-design.

As a comprehensive re-design and manufacture of the actuator was not possible within the original project timeframe, a simplified actuator was proposed that would fulfil the Südmo DN80 retro-fit requirement for the site trial, albeit with reduced functionality.

The simplified design was formulated at a review meeting between AAU, BHR and KM, and incorporated lessons learnt from the first prototype. Internal control valves were removed and the cylinder used screwed end caps to eliminate problems of welded construction. Generous seal lands were used to ease assembly and to prevent seal damage during assembly which had plagued the early design.
The Intellitop was mounted directly to the top of the actuator and was actuated by an extension shaft through the cylinder cap. The actuator mounted to the valve using the standard Südmo adaptor with a return spring to keep the upper seat lift valve closed.

1.3.3.4 Multi-Function/Regulating Actuator Design

A new specification for the HYDRACT actuator was drawn up based upon the following revised requirements from the end-user:

* The latest commercially available sanitary valve designs have additional features that allow for both upper and lower seat lift functions
* Revised seat lift force can be 10kN, which is over twice that originally expected.
* Interfacing with the plant controller is a high cost activity and the customer will want to re-use existing hardware if at all possible.
* Physical size is not an overriding consideration
* A regulating capability would greatly increase the flexibility of the valve

Successful site trials with the simplified HYDRACT actuator contributed to securing a 6-month project extension for a second site trial with a multi-function/regulating actuator.
Design review meetings were held to discuss the options for the new actuator design, and the choice was limited by the allowable timeframe.

A compromised design contained the following features;

* Stainless steel tube cylinder
* Commercially available control valves at hydraulic power unit
* Return springs on seat lift pistons
* External position sensor
* Bi-directional seals

This would fulfil the functional requirements for the site trials without being overly complex or requiring specialised miniature hydraulic components.
Manufacturing drawings for the new actuator were produced by KM who also machined the components.

1.3.3.5 Production Actuator Designs

The multi-function/regulating actuator design of 1.3.3.4 fulfilled all the operational requirements for the present sanitary valve market. However, the capital cost of the present HYDRACT valve actuator system is likely to be higher than commercially available pneumatic alternatives. Cheaper running costs and savings in plant layout due to the regulating capability of the actuator go some way to offset the higher initial cost, but this will not be obvious most end-users.
For this reason it is desirable for the production actuator system to minimise high cost bought in items, such as hydraulic control valves, and further development is needed in this area. The original fully hydraulic prototype actuator with internal control valves still has potential as the basis for a production unit.
A ¼-turn actuator solution is also needed to meet the demand for a complete HYDRACT solution for the sanitary valve market, and some novel concepts are available for consideration.

1.3.4 Seal Development and Test

LUH

1.3.5 Control and Communication

UM

1.3.6 Integration and Testing

The integration and testing phase of the project involved the production of a prototype HYDRACT system and a laboratory test rig, followed by performance/endurance testing of the actuator.

1.3.6 1 Design and Construction of Facility for Dynamic Testing of Actuator components

The hydraulic power unit described in 1.3.2 was designed to be used for both laboratory testing and the later site trials. The hydraulic system components were outsourced, but the system assembly and fabrication of the framework were done in-house at BHR Group Limited (figure. 1).

A performance/endurance test rig was required in order to obtain data from the actuator in terms of force, stroke, leakage, power consumption, maintainability and reliability.
The general concept was to mount the actuator to a rigid frame and apply external loads to mimic the forces likely to be experienced in service. An oil hydraulic system was chosen for its simplicity and because suitable donor components could be sourced local to BHR Group Limited.
A general view of the test rig is shown in figure 2, with the test actuator mounted on the right-hand side, and the two load actuators mounted on the left-hand side.

A 3-phase, 5.5kW motor powered a double gear pump that allowed independent hydraulic circuits to run the open/close and flush load cylinders. With a pressure rating of up to 275 bar, the rig had the capability of exceeding the loading requirements of the actuator test programme, and hence care was needed in its operation.
The main instrumentation on the test rig comprised of a 25kN bi-directional load cell, and two independent sensors for measuring the main actuator rod and flush shaft positions. The test actuator was fitted with two hydraulic pressure transducers and thermocouple to respectively measure the line pressures and main body temperature. Any external leakage from the actuator seals could be measured using a simple funnel and measuring cylinder arrangement that had to be manually recorded at relevant points in the test programme.

A PC based data acquisition and control system was custom built for the HYDRACT project using National Instruments hardware and LABVIEW software.
During the laboratory trials at BHR Group Limited the PC took the place of a plant controller and signalled the HPU to open/close/flush the HYDRACT test actuator. This could either be done manually (by clicking icons on the computer screen), or automatically (by running a cycling program) when endurance testing was required. The PC also controlled the performance/endurance rig, operating the direction and sequence of the loading rams to coincide with the actuator function.

1.3.6.2 Actuator Assembly and Commissioning
Assembly of the first prototype actuator required specialised tooling to minimise seal damage during installation. Re-machining was necessary on some components to improve the seal lead-in and this was followed by polishing. Surface finish measurements of the main cylinder bore were acceptable over the main stroke area, but the flush piston bore was not ideal.

The internal valves used hardened stainless steel balls and springs against polished stainless steel seats, but were prone to leakage under test. The next step was to make miniature stainless steel poppet valves that could be lapped in using a custom jig. These were also prone to leakage in a short time period, and close examination revealed erosion of the seating area.
Both PTFE and PEEK poppet valves were made and tested, but both were prone to sticking and leakage.

The decision was made to use the original concept of hardened steel balls and polished conical seats for the prototype, as the balls were commercially available and the leakage could be measured and allowed for in the performance tests.

1.3.6.3 Performance Tests

The first prototype actuator was subjected to load tests at 100bar operating pressure and resulted in opening/closing forces of 12.6kN and 13kN respectively. The maximum flush force measured was 1.85kN but the flush operation proved to be unreliable due to internal leakage problems with the actuator.

1.3.6.4 Endurance Tests

The endurance tests were designed to highlight any potential problems that may arise during the proposed 90-day site trials, bearing deliverable deadlines and the need use the existing prototypes for the later trials.
Two main tests were deemed necessary from a design of experiments, one using tap water at room temperature, and the other at an elevated temperature of 80°C. Tap water was likely to be the worst case in terms of lubrication, friction, wear and leakage and hence was expected to be a harsh test. Operating at the high operating temperature of 80°C was likely to exacerbate the situation and was the limit of the pressure transducers used on.
Running each endurance test for 1000 cycles seemed to be a sensible compromise between the limited cycles envisaged on site, the need to ensure an adequate life of the test component, and the testing time available.
Both tests were performed at a nominal operating pressure of 50bar to open and close the actuator, as this had proven to provide a usable flush force when the full system pressure of 140bar was applied in the closed position.
The room temperature endurance test resulted in several mechanical failures of the main piston shaft and this led to a revision of the design. It was also obvious that the welded end cap on the cylinder had distorted the cylinder bore, thus causing the piston to stick. This problem was eliminated in future designs by threading the cylinder ends to ensure uniform tolerance along the bore.

The test actuator was repaired and rebuilt for the hot endurance tests where it was placed within an environmental chamber on the performance test rig. 1000 cycles were completed without mechanical failure, although there was a slight water leak at the main shaft seal. The post test inspection revealed some wear on the main cylinder/seals and a seized flush piston.
The system was sterilised prior to the hot endurance test and remained sterile for the duration of the test, thus confirming the sterilisation procedures and equipment.

1.3.6.5 Sensor and RFTB Integration

The sensor and RFBT prepared by University of Minho/Norsisemas were integrated with the HYDRACT system and tested at the laboratories of BHR Group Limited. BHR mounted the receiver and relays within the main control cabinet of the HPU, while the transmitter was mounted on or near the actuator. For the trials the PC was wired to the transmitter to simulate the plant controller commands. The commands were then sent by RF to the HPU to perform the necessary open/close/flush operations.
UM performed signal strength tests while at BHR and had no problems with signal transmission across the laboratory. The RFBT was used throughout performance and endurance testing with no indication of a system fault.

UM performed tests with various magnets and sensors within a KM supplied cylinder, proving that a Balliff sensor and Eclipse magnet would give adequate resolution. When the set up was transferred to the prototype actuator there was a significant reduction in the signal resolution due to differences in the cylinder material. Because of the conflicting requirements between the sensor and choice of cylinder material, the sensor development was put on hold. This part of the work was subsequently overshadowed by the end-user requirement to use the existing valve position sensor in a retro-fit application.

1.3.7 Installation and Operation of Site Located Valves

1.3.7.1 Site Trial Preparation

A meeting was held with Carlsberg Production Management and Plant Engineers in January 2010 to discuss and plan the site trials scheduled for early that year. It immediately became obvious that Carlsberg's specification for the trial actuator had changed from that discussed in November 2008, and that a significant redesign was required.
The new specification called for a retro-fit to a Südmo DN80/100 type 620 valve that has multiple seat lift functions and had to retain the 'Intellitop' communication system with the Siemens PC7 plant controller. The rationale behind this was to provide a cost effective solution to replacing existing valve actuators within the plant that are very costly to repair. There is also a high labour cost associated with re-programming new controllers into the plant, hence the desire to re-use the existing valve tops with their associated position sensors.
BHR, KM and AAU reviewed the options and with Carlsberg's approval a simplified actuator was designed and built for the trials as a retro-fit to a Südmo DN80 valve.
The new actuator was manufactured by KM and trial assembled in Denmark before being shipped to BHR in England, along with a Type 8680 Intellitop and a DN80 valve body.
BHR performed the final assembly with GFD supplied seals, modified the Intellitop for RF communication with the hydraulic power unit, and performed operational tests with a field data acquisition system.
The hydraulic power unit was modified to operate at 100bar with a corresponding change to the accumulator pre-charge pressure. A series of actuator cycles were performed, controlled through the Intellitop and RF communication network, to give confidence in the system operation.
The base-line data showed a peak power consumption of 1.8kW and a stand-by power consumption of 77W, mainly associated with 24VDC solenoids on the control valves. A rationalisation of the power characteristic yielded an energy consumption of 525J per stroke, compared with a theoretical hydraulic work of 510J to stroke the actuator.

1.3.7.2 Plant Installation

The HYDRACT water hydraulic power unit, trial actuator and portable data acquisition system were transported by ferry/van from BHR Group Limited, England to the Carlsberg brewery in Denmark, mid-May 2010.
The site chosen for the demonstration was a clean-in-place plant that was conveniently inactive on the day of installation. Figure 3 shows a general view of the plant.

The hydraulic power unit is compact and could easily be sited next to the plant without causing an obstruction to plant personnel during normal plant operation/maintenance or to on-going plant construction.

The hydraulic installation was completed during the first morning of the trials and the system set up to operate in a manual mode (without the plant controller) during that afternoon. A demonstration was performed for Carlsberg engineers with the actuator on the plant valve, operating in manual mode from the HYDRACT PC. The closing speed of the valve was considered to be too fast, and so a restrictor was placed in the return line to ease the situation. This highlighted the need for additional flow control in later installations, and was accomplished with an adjustable flow control valve.
The following day, the HYDRACT system was run from the main plant controller as part of the normal CIP programme. This took several hours and the data acquisition system was used to monitor the system performance for comparison with the pre-site trials. Carlsberg staff reported no operational problems from the valve and so the HYDRACT actuator was left to run unattended during the rest of the CIP cycle.
A dip slide was taken from the HPU water tank after installation and incubated by Carlsberg staff in their laboratory. The indications are that a moderate growth of bacteria/yeast was evident on the slide, which was not surprising considering that the actuator had hoses had been disconnected during installation. This highlighted the need for re-sterilising the system following a break into the system,
Confident that the equipment was performing satisfactorily, the HYDRACT system was left in the plant for further evaluation.
Radio frequency based technology (RFBT) was used as the sole communication route between the Carlsberg plant and the HPU, and performed faultlessly throughout the plant trials. Although the transmission distance was relatively short at the Carlsberg site, there is every indication that the technology is sound for future implementation.
The demonstrator was removed from the CIP plant on 26th May 2010 by KM staff, having completed two weeks of trouble free operation. The only slight criticism by Carlsberg was that the valve was still closing too quickly, but other than that they were very happy with the way the HYDRACT demonstrator had performed.

1.3.7.3 Energy Comparison

It was not possible to take energy consumption measurements for pneumatic valves while at the Carlsberg site and so a unit was taken back to BHR Group for laboratory measurements. The main stroke of the actuator was connected to a mini air compressor and the air flow measured to the compressor by a hot wire anemometer. The average air consumption was found to be 8.26 litres when compressed to 7 bar, or 1.72kJ per valve stroke. Comparing this with the HYDRACT actuator gave the water hydraulic option a 70% energy saving based upon the rationalised power measurements and a spring return. With a hydraulic closing function the HYDRACT actuator will still give a 39% energy saving compared with the spring return pneumatic valve actuator.
Laboratory and plant measurements thus suggest that the theoretical energy saving of 50% is realistic for water hydraulic systems when compared with the compressed air alternative.
However, a significant additional saving can be realised by the virtual elimination of wasted energy through compressed air leaks that contribute to an astonishing 30-50% of the annual energy cost. These figures are typical of compressed air installations throughout Europe and were confirmed during a recent seminar at the 2010 IFPEX Exhibition 'Compressed Air - The Importance of leak detection' by Chris Dee, Executive Director, BCAS (British Compressed Air Society).
Based upon these figures, the water hydraulic option has the potential to cut the energy bill by at least 65%.
The annual energy bill at the Frederica plant is around 11M€, but only a portion of this is for compressed air as there are many pumps and heaters throughout the plant. The main compressor house contains three 250kW compressors, and they all run during normal production, 5-days a week. One 250kW compressor runs when the plant is not producing (at the weekend) as it is required to keep the system pressurised while compensating for air leaks throughout the system.
To calculate the annual cost associated with the air supply for the plant we can thus assume one compressor runs 24/7, while the other two run at 50% duty 24/5.
Assuming 4.5p/kWh would give an annual energy bill for compressed air of around £168,000 or 200,000€. Switching to a water hydraulic solution could save 130,000€/year.

1.3.7.3 Second Plant Trials

In December 2010, BHR and KM were back at Carlsberg, Fredericia, for a second series of plant trials with a new multi-function/regulating valve actuator.
The second demonstration required an extension to the project and involved a re-design of the actuator and HPU. The new actuator could perform both upper and lower seat lift functions, and featured closed loop position control of the main open/closed valve stroke. The main objective of this demonstration was to prove the additional functionality of the new design and the potential benefits that smart hydraulic actuators can bring to plant layout and control.

The trials were performed at the same location as the previous trials with the system commissioned off the plant, figure 4. A demonstration allowed Carlsberg to witness the multi-function and regulating capabilities of the valve with easy viewing of the internal components, which was not possible when installed within the plant. The demonstration also allowed the system to be sterilised prior plant installation, which was important if there were any leaks from the hydraulic system that could indirectly contaminate the process fluid. A dip slide was taken from the HPU tank after sterilization and showed no bacteria/yeast or mould after 48 hours.

Carlsberg made several observations from the multi-function demonstration that resulted in adjustment to the upper/lower seal lift strokes and a further slowing of the operating speed.
These modifications were demonstrated to Carlsberg and they were satisfied with the set-up, giving approval for the valve to be installed within the CIP plant.
Having successfully proven the multi-function aspects of the new actuator, the control system was switched to closed-loop position control and a demonstration given of the actuator's ability to hold the valve in any desired position from fully-open to fully-closed.
The ability to control the speed of operation without the manual flow control valves was also demonstrated, and Carlsberg were obviously impressed with the possibilities that this offered. Unfortunately it was not possible to drive the proportional valve through the closed loop controller while under Carlsberg plant control, and hence the speed would have to be set by the manual flow control valves for the in-plant demonstration.

Installing the valve/actuator into the CIP plant involved shutting the manual isolation valves and disconnecting the air supply from the local double seat valves (effectively ensuring that they stay in the closed position). The original valve/pneumatic actuator were then removed from the pipework and the HYDRACT valve/water actuator installed in place. The Intellitop from the plant valve was swapped onto the HYDRACT actuator (with a modified cap) and wired up to the hydraulic power unit. Finally, the position sensor was mounted on the new Intellitop and the connection made to the main plant controller. The HYDRACT actuator was run manually to check operation before the air supply to the local valves was reinstated the isolation valves re-opened.

The plant control room was left to run the valve for an hour under their normal procedure and reported no indicated problems with the valve.

System operating pressures were recorded with the valve in the plant, and showed a slight increase in the pressure required to move the valve from 7 bar to around 10 bar.

The power consumption measurements were seen to be similar to those taken previously, with a peak power demand of 1.03 kW and 47W on standby.

Carlsberg engineers considered it unnecessary to perform regulating trials in the plant as they were happy with the off-plant trials, and could not measure the effectiveness of a regulating valve at that location.

Potential Impact:
Potential impact
The potential economic impact on the consortium was already from the beginning of the project in 2008 considered to be significant. It would generate a profitable business by manufacture and sale of actuators to valve manufacturers and end users as well as actuators for retrofitting onto existing valves. The consortium also expected a considerable income from selling licences to existing valve manufacturers; this would allow the customers to choose between buying complete actuators from KM Rustfri or to manufacture themselves.

KM Rustfri is planning to build up a supply chain and a sales organization to manufacture and sell actuators both for retrofitting and new installations on behalf of the consortium. KM Rustfri is also planning to offer complete retrofit installations on the end users site.

The expected benefits for the customers arriving from reduced energy consumption by ap. 70%, a similar reduction in CO2 emissions and reduced waste is expected to offer a payback time for the end users from 1 to 3 years and hence generating a profitable business for the consortium.

The potential turnover for the consortium is estimated to be 85 m€/ 3 years after launch with a profit before tax above 10%.

When the project was launched the primary focus was on energy savings, reductions of CO2 emissions, increased reliability and reduced waste. As the project progressed other advantages such as reduced capital investment in new installations were better understood. The improved precision and reliability over pneumatic actuators and also the fact that even existing valves can be used as regulation valves has revealed a potential benefit for the end user that may even exceed the benefit from reduction of running costs. The end user will be allowed to re-think all future installations and thereby reducing the part of the investment in new facilities related to pipe, valves and pumps between 25% and 40%.

The potential socio-economic impact is difficult to estimate in precise terms and numbers, as the potential for the industry by embracing the new technology is growing as the development work is progressing.

Dissemination activities
The original dissemination strategy was based on an open approach towards the sanitary industry with the ambition to involve end users and valve manufacturers as well as universities, scientific magazines, internet, project homepage, newspapers, NGO's etc. The knowhow would furthermore be disseminated via the publication of the patent application.

The consortium anticipated that a fundamental change of technology from pneumatics to water hydraulics would be met with both scepticism and hesitation by the industry. It was therefore considered essential to have the support from a global end user based on a successful test before intensifying the dissemination activities.

In the summer of 2009 the consortium decided to present the concept to leading valve manufacturers who all acknowledged the concepts potential. Unfortunately some of the reactions were directly hostile and threatening towards the leading SME KM Rustfri. It was then decided to take a step back and chose another strategy involving intensive dissemination activities at the time when the actuator was launched on the market.

The dissemination activities will therefore be integrated with the plan for exploitation and the first major activity will be the Brau Beviale Tradeshow in Nuremberg Germany in November 2011.

Exploitation activities
The fact that water hydraulics technology is completely new to the sanitary industry has been given a lot of consideration and the consortium has been determined not to underestimate the natural scepticism and hesitation of an industry that has been using pneumatic systems for at least the last 4 decades.

The exploitation strategy has therefore remained unaltered since the outset of the project. The approval of a global end user who can influence peers both in the same industry and in related industries is seen as the key to obtaining the necessary impact in our dissemination and exploitation activities.

The leading global technical tradeshow for the brewery industry is Brau Beviale on 9-11/11-2011 in Nuremberg.. Brau Beviale has been chosen as the best place to launch the product because a brewery tradeshow will give the consortium maximum benefit from having Carlsberg as end user.

The next tradeshow will be Anuga Food Tech in Koeln in February 2012 which is the largest tradeshow of the year with focus on the dairy industry. The 3 months gab between the two tradeshows is ideal, because it will give the consortium time for preparation including a test on a suitable dairy installation.

Achema in Frankfurt 2012 has been chosen as the place to launch Hydract in the pharmaceutical industry. Once again a 4 months gab from the previous tradeshow is ideal for preparation and field tests.

Finally the original consortium consisting of KM, NS, GFD and also BHR who in this context will be an SME has applied for support under the FP7 Demonstration action to intensify the demonstration activities as one of the most important exploitation initiatives.

Conclusion
Since the project was started in 2008 and following the increasing awareness on CO2 reductions following the financial crisis in 2009 to 2010.

The sanitary industry covering the food industry and the pharmaceutical industry can benefit from reduced energy costs, reduced waste and reduced investment in new installations. The immediate benefit for the end users is the described savings on running costs but water hydraulics will furthermore enable them to fulfil the still increasing targets for reduction of CO2 emissions. As the confidence and experience in water hydraulics grows and the end users learn to appreciate the increased precision and reliability of a hydraulic system over pneumatics, the end users can start to design installations differently and thereby reduce the investment in new facilities.
List of Websites:
www.hydract.dk

Contact details, see attached document

Final Report Summary - HYDRACT (Hydraulic Actuator for Valves in Brewery, Dairy and Pharmaceutical Industries)

Share this page

Download