Docstoc

JISC final report template

Document Sample
JISC final report template Powered By Docstoc
					Project Acronym: ECCILES (Energy Conservation in Computer Intensive Environments)
Version:1
Contact : Derek Rout
Date:30/4/10




Bolton ECCILES Final Report



                                                   Project Information
Project Acronym                     ECCILES
Project Title                       Energy Conservation in Computer Intensive Environments
Start Date                          Oct 2008                       End Date           June 2010
Lead Institution                    University of Bolton
Project Director                    Derek Rout
Project Manager &                   d.rout@bolton.ac.uk
contact details                     01204 903014
Partner Institutions
Project Web URL                     http://www.bolton.ac.uk/projects/JISC/ECCILES
Programme Name (and                 Synthesis and Benefits Realisation; Institutional Innovation
number)
Programme Manager                   Lawrie Phipps




                                                    Document Name
Document Title                      Final Report
Reporting Period                    Oct 2008 to June 2010
Author(s) & project role            Derek Rout – Project Director
                                    Darrell Slater-Smith – Project Manager
                                    Patrick O‟Reilly – Head of Information Systems and Technology
                                    Danny Morton – TRIZ support
                                    Roger Kirkman – IT technician
                                    Mark Williamson – Network and systems Manager
                                    Rina Lakhman – Project administrator

Date                                31 May 2010              Filename
URL
Access                               X Project and JISC internal                General dissemination

                                                   Document History
    Version                   Date                                       Comments
1                      31/05/2010              First draft
2                      31/08/2010              Read and revised by project director




Page 1 of 37
Document title: JISC Final Report
Last updated: May 2010
Project Acronym: ECCILES (Energy Conservation in Computer Intensive Learning EnvironmentS)
Version:1
Contact: Derek Rout
Date: 30 April 2010




JISC Final Report




     Energy Conservation in Computer Intensive Learning
                       EnvironmentS




Patrick O‟Reilly
Derek Rout
Mark Williamson
Danny Morton
Rina Lakhman
Roger Kirkman
Darrell Slater-Smith




                                                   Page 2 of 37
Project Acronym: ECCILES (Energy Conservation in Computer Intensive Learning EnvironmentS)
Version:1
Contact: Derek Rout
Date: 30 April 2010




Table of Contents

Contents

  Acknowledgements ............................................................................................................................. 4
  Executive Summary............................................................................................................................. 5
  Background ......................................................................................................................................... 6
  Aims and Objectives ............................................................................................................................ 6
  Methodology ........................................................................................................................................ 7
     Each solution generated and adopted by the project team has been turned into small case
     studies of investigation..................................................................................................................... 8
     Overall methodology applied ........................................................................................................... 8
     Each case study used a methodology which best suites the case being investigated. .................. 8
  Implementation .................................................................................................................................... 9
     The Project Team ............................................................................................................................ 9
     Working Group ............................................................................................................................... 10
     Project Case Studies ..................................................................................................................... 11
     1.      TRIZ Case Study ................................................................................................................... 12
     2.      Metering Case Study ............................................................................................................. 13
     3.      Lighting Case Study .............................................................................................................. 14
     4.      Energy Modelling Tool .......................................................................................................... 14
     5.      Ventilation System Case Study ............................................................................................. 16
     6.      Awareness Case Study ......................................................................................................... 18
     7.      WAKE on LAN Capability ...................................................................................................... 18
     8.      Survey of Lab Requirements and Student Demands ........................................................... 19
     9.      Powerman Case Study ......................................................................................................... 19
     10.         Virtualisation ...................................................................................................................... 20
     11.         Server Room Temperature ............................................................................................... 20
     12.         Embedding ........................................................................................................................ 21
  Outputs and Results .......................................................................................................................... 23
  Outcomes .......................................................................................................................................... 28
  Conclusions ....................................................................................................................................... 33
  Implications ........................................................................................................................................ 35
  Recommendations (optional)............................................................................................................. 36
  References ........................................................................................................................................ 37
  Appendixes (optional) ........................................................................................................................ 37




                                                                  Page 3 of 37
Project Acronym: ECCILES (Energy Conservation in Computer Intensive Learning EnvironmentS)
Version:1
Contact: Derek Rout
Date: 30 April 2010



Acknowledgements
The ECCILES team would like to take this opportunity to thank the JISC Institutional Innovation
Programme for making this project possible. Thank you to the entire team at JISC for all your hard
work and support for the ECCILES team throughout the project and with establishing possible links to
the wider JISC community. Their knowledge and access to various projects in the sector helped us to
connect our project with others.

Without the funding, guidance and support from JISC this project would not have been able to achieve
the collaborative solutions delivered by the team. Their enthusiasm to see how TRIZ would be
applied to the problem has enabled the team to dig deeper into energy saving measures and come up
with both technical and operational solutions.

Other groups to acknowledge:

The Carbon Trust for their contribution to the awareness training programme.

As project leader I would like to express my joy at working with such an enthusiastic dedicated and
flexible team. So often groups become talking shops but this group due to hard work and
commitment has achieved a great deal.

The project has brought together a team with diverse backgrounds and skills which has stimulated
debate and better understanding not only of the challenges of their own but in wider areas.

I would like to name the members of the ECCILES team who have contributed so greatly to the
success of the project Darrell Slater-Smith, Patrick O‟Reilly, Danny Morton, Roger Kirkham, Mark
Williamson, Rina Lakhman and Craig Charlton.


Pre-note
This document is layered; the different achievements of the project are further explained in case
studies of their own. This document has a summarised overview of each “product”/ case study and of
the overall approach for the project. There are links in the document to the relevant case studies
which further detail the approaches taken and the results from the projects undertaken.




                                                   Page 4 of 37
Project Acronym: ECCILES (Energy Conservation in Computer Intensive Learning EnvironmentS)
Version:1
Contact: Derek Rout
Date: 30 April 2010


Executive Summary
Computers have become the essential tools in many design, engineering and creative media
disciplines found in further and higher education. Specialist hardware and software and their new
capabilities, have provided a significant driver for curriculum resource planning and development at
the University of Bolton, and has led to a significant increase in computing capacity and power. The
number of specialist computer laboratories has increased to support computer games development,
special effects, animation as well as general computing and networking courses.

The JISC ECCILES project focuses on reducing energy, and consequently carbon emissions, due to
these computer learning environments which need greater processing power and data storage to deal
with new digital techniques e.g. digital image and video manipulation. The ECCILES project has
taken a holistic view of all the energy issues from the technology, the physical classroom
environments and the back-end supporting infrastructure such as servers and server rooms to
develop solutions to reduce energy consumption.

In these curriculum areas, resource planning has, traditionally, been demand led. The focus has been
fully on investing to support curriculum development using the latest tools: issues of sustainability,
including environmental, have not been major concerns.

The ECCILES projects worked with multiple stakeholder groups using the TRIZ innovation
methodology, a process of systematic innovation, to investigate the systemic and technology
problems that lead to excess energy consumption in these learning environments. The TRIZ tools
were used to develop different perspectives on the problem including its context, the systems
excesses and deficits, and what people say about the problem. The TRIZ 40 inventive principles for
problem solving were selected and applied to generate a range of solutions. In particular, the
stakeholder groups which were (IT, estates, academic and student), found the use of perception
mapping a particularly powerful tool for negotiating a common understanding, and is the basis for
generating a range of operational, technical and organisational solutions.

The resulting solutions included:
    Finer grain measurement and feedback of energy usage information
    Improvements to the physical environment in classrooms – lighting, cooling, air flows
    Hibernation of idling computers
    Optimization of timetabled computer rooms and drop-in facilities
    Free air cooling systems in classrooms and server rooms
    Server virtualization

The project developed a more detailed understanding of the wider issue of how current curriculum
resource planning and development is driving energy consumption upwards and has put forward
proposals of how the University might improve the management of this.




                                                   Page 5 of 37
Project Acronym: ECCILES (Energy Conservation in Computer Intensive Learning EnvironmentS)
Version:1
Contact: Derek Rout
Date: 30 April 2010


Background
In 2008, like all Universities, Bolton has experienced up to 73% increase in energy bills over recent
years and was anticipating another 100% increase. Managing down energy usage is a business and
ethical imperative. Previous efforts, focused on space and resources, including consolidation onto one
campus to reduce costs.

Beyond small scale applications of green initiatives, ”Switch it Off” campaigns etc, the University had
neither the capacity or resources to assess which green computing technologies and approaches
would give the greatest benefit. Modest savings in one area were often wiped out by growth in
consumption in another. Computer intensive disciplines continue to demand increasing processing
power with consequent heat generation and increasing electricity consumption.

As an exemplar case study the University decided to focus on one of its major sites for the ECCILES
project. The Deane site houses the School of Business, Computing and Creative Technology, and the
School of Built Environment and Engineering. At the start of the project, the Deane campus had 350
student workstations located in 14 computer rooms, specialist laboratories, a 120 machine server
farm and a medium size data centre housing 36 servers including a SAN, and thin client clusters. In
previous years sections of the site had been refurbished and air conditioning had been retrofitted.
However, the majority of computer teaching rooms are located in older areas, and due to the
construction of the building temperatures regularly exceed 28°C. These rooms contain high end
workstation used for computer games development, digital rendering, video and special effects work.
There was considerable pressure from staff and students to install air-conditioning to achieve a more
comfortable learning environment.

Overall, whilst the computing resources had grown significantly, little attention had been given to
optimising energy use. Issues of sustainability were concerned with developing new curriculum and
developing resources to support the curriculum activity. Energy management in this context, was
regarded as an encumbrance.


Aims and Objectives
The ECCILES project aimed to reduce the power consumption arising from computing intensive
teaching and learning spaces, whilst improving the environment for staff and students at the Deane
Site of the University.

The specific aims were:
1. To constrain, reduce the growth of, or reduce power consumption, directly and in-directly, in the
    use of computers at the Deane campus.
2.   To develop a deeper understanding of the factors (going beyond the symptoms) that encouraged
     the growth in power consumption around computing environments at Deane.
3.   To explore and demonstrate the practical use of TRIZ as a structured methodology for tackling
     complex technical problems and innovate in Universities, feeding this back to the JISC innovation
     programme.
The project objectives were:

1. To establish means of effectively measuring improvements in consumption or potential
   consumption of electricity at Deane, by establishing finer grain systems that monitor power
   consumption to areas of activity e.g. computer classrooms, computer server rooms, staff personal
   computers.

2. To establish a base line for power consumption at Deane from which improvements can be
   measured.

3. To build a multi disciplined team capable of breaking down problems and identifying solutions
   using the TRIZ methodology




                                                   Page 6 of 37
Project Acronym: ECCILES (Energy Conservation in Computer Intensive Learning EnvironmentS)
Version:1
Contact: Derek Rout
Date: 30 April 2010

4. To use TRIZ methodology to produce a comprehensive assessment of the problems and factors
   influencing energy consumption at Deane, looking beyond pure technology issues to pedagogical
   practice and administrative practice.

5. To use TRIZ to generate a wide range of possible solutions taking into account the resources
   available at Deane

6. To select and test energy reduction solutions and approaches, measuring their impact on
   consumption

7. To investigate the potential of Free air systems to reduce the need for cooling in computer rooms

Methodology
The project relies on engaging with the whole Deane community in being aware of, and
understanding, the energy issues and making energy conservation a more important a factor when
setting out planning strategies. The project team wanted to change attitudes and behaviours of staff
to adopt a greener way of working.

The project has used the TRIZ innovation methodology as the means of developing a deeper
understanding of the problem domain, of investigating the systemic issues that lead to excessive
energy consumption, and as the means of generating and selecting a range of solutions. Where
traditional brainstorming techniques can be sometimes be un-focussed, TRIZ provides a range of
tools to enable a more systematic approach to be taken. At the beginning of the project it was clear
that different stakeholders had different views on the origins of the problem i.e. the things that were
leading to access energy consumption and clear ideas on how they thought problems could be
solved. However, there was no consensus. One person‟s control becomes another‟s barrier to
productivity. This is typical for problem solving: improving one aspect of a problem (what TRIZ refers
to as reducing a “harm”) can give rise to another. There may be a tension or conflict between factors:
as one improves, another worsens.

Using the TRIZ tools with stakeholder groups, the project built up an in-depth and sophisticated view
of the way in which energy consumption has grown and factors that influence future energy
consumption. In summary, the project group used the TRIZ tools to model:

    1. The overall system in which the energy consumption problem associated with computer labs
       exists. The higher order system being the reasons the computer- dependent higher education
       exists. The subsystem issues being the efficiency and effective management of energy within
       the technology.

    2. How time effects energy consumption. The past, present, and future of energy consuming
       associated with computers in teaching. Here is possible to examine the types of activities and
       the reasons for their existence, but also the technology evolution itself from mainframe to
       cloud.

    3. The overall energy management system around computer-based curriculum and the
       excesses and deficiencies within this system e.g. computers that consume energy when not
       processing, students who can‟t get access to a computer when they need one.

    4. The perceptions of stakeholder groups and the impact these have on decision-making e.g.
       the decisions to increase resources.

The ECCILES project stakeholder groups met weekly throughout the project to work through each of
the TRIZ tools. Using structured tools improved the shared and negotiated understanding between the
different stakeholders groups. Applying the tools for generating solutions based on TRIZ‟s forty
inventive principles, ensured there was continuity from analysis to solution, and encouraged the team
to use a wider range of approaches. A full description of the application of TRIZ within the project is
set out in the ”TRIZ case study” (click to open) (Appendix 1)




                                                   Page 7 of 37
         Project Acronym: ECCILES (Energy Conservation in Computer Intensive Learning EnvironmentS)
         Version:1
         Contact: Derek Rout
         Date: 30 April 2010

         Each solution generated and adopted by the project team has been turned into small case studies of
         investigation.



         Overall methodology applied                  TRIZ Theory of inventive
                                                                                                   Overlying methodology used to
                                                                                                   break the problem down into
         To the problem                                   problem solving                          smaller manageable chunks.

                                                            J           A       M       P
Perceptions – broken down chunks                       K        D
                                                                            B       N       Q
of the project which can be solved                          E
                                                  L                 G           C           R
by one of the case studies                                      H                   O
                                                        I                   F




                Case study           Case study        Case study               Case study      Case study       Case study




         Each case study used a methodology which best suites the case being investigated.




                                                                    Page 8 of 37
Project Acronym: ECCILES (Energy Conservation in Computer Intensive Learning EnvironmentS)
Version:1
Contact: Derek Rout
Date: 30 April 2010


Implementation
The Ecciles team focused measurement and analyses on the one block of the Dean Campus “C-
block” rolling out generic approaches to energy saving measures across the wider university.
The Ecciles project worked with two different types of solutions:

Technical solutions – these usually require investment, in technology which are greener, e.g. free air
cooling, power management software, Fit PCs...etc. For all technical investments the project team
followed a few rules:
      Payback period – if the savings made did not cover the cost of the outlay within 5 years it was
        not worth investment at this stage in time, and would be considered again when the
        technology has improved, or when that room is removed..
      Is it easily installed and run?
      Will we need a dedicated resource to maintain the system?
      What are the benefits to the university?
      Is it future proof?
      There is no point in installing something when a better solution is around the corner.

Operational/ Organisational Solutions – By changing behaviours and working habits of campus
users, the project can contribute operational change university wide. Changes such as procurement
policies, awareness training, inventive use of facilities, green timetabling,...etc. The operational
solutions which the project team could implement have been implemented or are in the process of
being implemented and those which need support from the university‟s management are in the
“Embedding case study”, which will be put forward to the management group.

The Project Team
As described in the project plan Ecciles aimed to setup as a multidiscipline team of Facilities staff, IT
and end user academics and technicians. The team met once a week to work in a series of TRIZ
workshops. The TRIZ workshops used both structured and unstructured activities. As actions were
realised in the workshops respective groups investigated further and reported back at the Ecciles
team meetings.

The project team was split into three main workgroups:




                                           Facilities          User
                                                              group



                                                        IT



Each group had different concerns, priorities and perceptions of the problem and had different ideas
as to how to solve the problem.

Facilities – As the energy budget holder, facilities aimed to get the consumption of energy down.
             Generally saving energy is perceived to be a facilities problem. When we talk about
             conserving energy, and reducing carbon it is perceived that it involves, installing solar
             panels and modern energy efficient heating, lighting and air conditioning.




                                                   Page 9 of 37
       Project Acronym: ECCILES (Energy Conservation in Computer Intensive Learning EnvironmentS)
       Version:1
       Contact: Derek Rout
       Date: 30 April 2010




                IT – In charge of, computing and network facilities across the university. IT are concerned
                     about the rate at which computers and servers have grown in numbers across campus.
                     It is perceived by academic and admin staff and some student groups that IT provisions
                     are available on demand. IT support teams recognise that there are limitations to their
                     capacity in being able to support and maintain increasing numbers of computers.

     User Group – End user involvement on the project got better buy-in on solutions from the end users.
                 The academic staff from C-Block are best placed to contribute perceptions related to the
                 way in which the space is used. Implementing any kind of change may affect the end
                 user. If this effect is negative it may lead to rejection of the solution provided, erasing all
                 the good efforts of the team. The user group bring an operational perspective to the
                 problem.

       Working Group
       The mix of disciplines was very important to the project delivery. It allowed each workgroup to
       understand the barriers faced by the other workgroups. For example:

                 Estates Perceptions                                              User Group Perception

Too many labs are open with just                                                     Labs have to be open to give
one or two students using the lab,                                                   students access to the specialist
we still need to heat, light and cool                                                software or equipment they need for
those rooms. Machines don‟t get                                                      their work. Each lab is different in
shut down overnight                                                                  setup so all have to be available.


          Why are there specialist labs?                              A Universal setup affects the
          Why can‟t we have universal setup                           efficiency of the software. Some
          on all the computers?                                       licenses are per machine and
                                                                      therefore very costly to put on all
                                                                      machines.


                                           Put universally set machines which
                                           support the majority of the pathways
                                           into the 365 room a 24/7 access
                                           service in the university.


       The above argument re-enforces the need for the different disciplines effectively working together to
       find solutions which do not create harm, cause conflicts, or make anything else worse. It would
       have been an option for facilities to simply lock all the rooms at 5pm. This would have made things
       more difficult for the academic staff and have come up against opposition.

       With academics and lab technicians involved in the project, implementing solutions is much easier.
       Sticking points and problems were debated early on, and a better solution emerged in the end.
       Having had the discussions in TRIZ workshops and ECCILES meetings the technician for the school
       also implemented energy saving techniques, within lab setups. This created a shift in procedural
       behaviour when implementing future changes to lab setups. Now energy is an underlying factor in the
       decision making as to how a piece of equipment or software is deployed. Academics can now direct
       students to the 365 rooms or remote connectivity after hours, instead of opening a lab.

       Technicians are now able to shut down labs appropriately at an earlier time of day, saving energy on
       lighting, heating and cooling of lab buildings. This coupled with the introduction of powerman, energy
       efficient lighting and free air cooling. The amount of energy wasted on the computer intensive
       environment has reduced considerably.




                                                         Page 10 of 37
      Project Acronym: ECCILES (Energy Conservation in Computer Intensive Learning EnvironmentS)
      Version:1
      Contact: Derek Rout
      Date: 30 April 2010

      Other technical improvements to services, such as Wake on LAN and remote connectivity have also
      improved student access to lab facilities from home. Allowing technicians to power manage remote
      connectivity to university facilities. Benefiting, services to international campuses.


      Project Case Studies
      Each case study in the diagram below became a mini project which carried out feasibility testing, pilot
      and roll-out. You can click on the diagram to open the relevant case study.

                                                                 TRIZ Support
           All

                                                          Why use the
            12. Embedding                                    TRIZ
     Carbon Linked Budgetting                            Approach
     Senior Level realisation of Energy                                                                       Facilities
       Waste                                                                          2. Metering
     Computer Levy for support, running                  1.TRIZ                     Case Study
       and disposal                                    approach to the
     Software Change Strategy                         problem Case
                                                                                                3. Lighting
     365 room Vs Labs open all hours                     Study
                                                                                                Case Study
     Future funding linked to
       environmental measures
                                                                                                    4. Energy Modelling
                                                                                                      Tool Case study



                                                                                                    5. Ventilation
                                                   ECCILES                                           Case Study
     11. Server Room
       Temperature



                                                                                              6. Awareness Case Study
                                                                                               Dinorwig
       10. Virtualisation                                                                      Poster Campaign
                                                                                               Senior Level Training
                                                                   7. Wake on                  Champions
                             9. Powerman                               LAN
                              Case Study                           Capability
                                              8. Survey of Lab
                                             Requirements and
                     IT                       student demand                                                         Overall




                                                      User Group




                                                        Page 11 of 37
Project Acronym: ECCILES (Energy Conservation in Computer Intensive Learning EnvironmentS)
Version:1
Contact: Derek Rout
Date: 30 April 2010




1. TRIZ Case Study

    TRIZ has been an effective approach to the problem. It‟s enabled members of the group to
    understand the problems faced by colleagues when changes are implemented.

    An overview of TRIZ:
             Problem
             Definition   Barriers to
                          solving the
                           problem        Broader
                                          Problem
                                                         Narrower
                                                         Problem
                                                                       Perception
                                                                        Mapping
                                                                                     Conflicts and
                                                                                     Contradictions


     The problem is defined by looking at:
     The Past – what has occurred to lead us the current problem;
     The Present – where are we now and why?
     The Future – will the infrastructure change and how will it impact on any solutions generated
     today.

     Super-system: Looking at the environment outside of the university (industry, legislation,
     external influences such as funding bodies influencing funding streams based on carbon
     reduction ability,…etc),
     System: Looking at the system within the university infrastructure. How and why the problem is
     the way it is, and avenues for improvement.
     Sub-system: Looking at area within the system, what is causing the problem? Is the solution
     local or does it involve influences from the above systems? Can we focus on individual rooms
     and solve the problem?

     A combination of the multi-disciplined team and the above focuses, help define the problem
     better, giving a more 3-dimentional view of the problem. It is easy to try and come up with the
     solutions, as a problem is highlighted, however as the problem is defined further, behaviours and
     perceptions affected the solutions and a more in-depth approach was required. For example
     initially the team sought to analyse the computer usage within the labs and find ways of cutting
     lab opening times or implement shut down strategies. By the end of the analysis process it was
     clear that a shut down strategy would hinder the business service provided by the university.
     This lead the team into looking at power management for PC‟s and Wake on LAN facilities, for
     remote connectivity.

     Perception mapping has been a particularly powerful tool for the ECCILES project. Sometimes
     projects are driven by the “loudest voice in the room”. With TRIZ all members of the team
     contributed their perceptions. All the perceptions are agreed by the workgroups, and the list is
     then mapped into a diagram shown below. The tool enabled the group to get a common
     understanding of the problem. The diagram aided a visual approach to the problem and how to
     start solving the problem. The team had to look at the effects and impacts a solution has on the
     systems defined above. “If by solving one problem, are we affecting or creating another?” The
     team had to address; What gets worse, and what is stopping us from solving that problem? This
     highlighted the easy fix solutions, which were implemented and the not so easy organisational
     and operational solutions which require more work and higher level management “buy in” in
     order to implementing a solution.




                                                    Page 12 of 37
Project Acronym: ECCILES (Energy Conservation in Computer Intensive Learning EnvironmentS)
Version:1
Contact: Derek Rout
Date: 30 April 2010




     The diagram below shows the user groups‟ perception map.

                                            ECCILES PROJECT
     V
                                        USER PERCEPTION MAP
          A
                         R                             S                       H         O
                  X
              U                                 Q

              B                        C                   I                 J
                  D
                                                E                                              P

                               F
      N
                                           G
                                                                                         L
                                                       Y
                                                                                                       PERCEPTION
                                                                           K                           LINK
                                         W
                                                                                                       PERCEPTION
                                                           Z                                           LINK/LOOP

                                                                                                   ?   PERCEPTION

                                            T                            M
                                                                                                       COLLECTOR
                                                                                                   ?
                                                                                                   J   POINT




     TRIZ and it‟s application for ECCILES goes into further detail in the TRIZ case study..

     Each letter in the diagram above relates to a perception which is listed in the TRIZ case study.
     The diagram demonstrates where the project team can look for the quick solutions. The red loop
     is a self-reinforcing loop, and by breaking just one of the links in the loop the team can address
     the rest of the perceptions in the usual way.


2. Metering Case Study
    Early on in the project, the team set about deciding the appropriate monitoring devices which
    needed to be installed. Given the age and design of the building the project team wanted to look
    at more than just power consumption. Academic staff expressed a need to improve the working
    conditions within the building. The heat within the rooms was an issue, with the location of the
    labs being in one of the hottest spots on campus plus heat generated from the computers plus,
    very low airflow through the block, the project needed to also, look for temperature and humidity
    monitoring.

    There are already companies who provide sub-metering facilities, and the installation was
    tendered for and installed. Once installed, the system provided the PTL (project Technical
    Leader) with the ability to measure power usage down to the date, time and segregate between
    lighting and computer power. As a university we could now analyse baseline energy consumption
    in more detail and target unexpected energy peaks for example on evenings and weekends,
    when consumption should be low.




                                                      Page 13 of 37
Project Acronym: ECCILES (Energy Conservation in Computer Intensive Learning EnvironmentS)
Version:1
Contact: Derek Rout
Date: 30 April 2010

    The metering case study details how the project team went about installing the sub-metering
    system, and shows the baseline data analysed. The data is continuously gathered and stored.
    Mining this data would give the university the ability to map activity to the energy consumed and
    possibly have a finer grained management of energy use.

    The sub-metering system has given the PTL a huge advantage in being able to segregate rooms
    and blocks in trying to understand where the energy goes within the building. This is the first time
    such data has been available to the university and the first time the PTL could focus on the peaks
    and understand what activity was consuming so much energy. The PTL also aimed to address
    the Baseline. Establishing what equipment was consuming energy continuously and if it was a
    necessary consumption of energy. Understandably this is a continual process as the equipment
    profile within the building changes from one year to the next. A full audit trail of the energy
    consuming equipment, and upgrades would help justify changes in energy consumption, and also
    advise schools on how to control these increases in baseline and average energy consumption.


3. Lighting Case Study
    As the project team debated about how energy savings could be made. Lighting was one of the
    obvious consumers of energy. C-Block was still lit by the old T12 and T8 lamps and switched
    start units which have an increased energy consumption. By replacing the lighting to T5 lamps
    can save the university 43% of power consumed on lighting.

    In addition to changing the lights, the light switches were replaced with Microwavable PIR sensors
    which detect body movement and heat in the room. The idea was to remove responsibility of
    turning lights off from the staff and student to an automated system. Some staff had the old
    perception of “the starting or striking of a lamp tube, consumes more electricity than if you simply
    left it on”, this was true for old lamps but no longer the case for the T5 lamps, eliminating the
    concerns that staff may have had.

    Human element – it was surprising how staff showed a dislike and resistance to the new
    automated lighting system. They could not come away from having a light switch to control the
    lights in the room. Some even went so far as to say that the lights gave them a headache. The
    PTL used a lux meter to dispel these perceptions. Some staff found it a good idea in order to
    save energy. Almost a year on from the installation of the lighting system, staff have now settled
    into the new system and have accepted it.

    One comment to note. Some rooms had a superior lux meter, with sensors on each light. The
    lights went from a bright window to a dark corridor and had an effective lux adjustment to suit the
    ambient light from the window. The sensors reacted very well with this setup, and the system
    was received well. However the rooms with only one sensor in the room were not as good.
    Lights were going out whilst exams were in progress simply due to the positioning of the sensors.
    When investing in a lighting system it is worth investing in the superior system to make it better
    received by the users. It is easy for end users to pass comment on failures of new systems and
    prove how the idea was not thought through and dismiss the new systems as a gimmick.

    Lighting Case Study highlights the detailed savings made by changing the lighting system.


4. Energy Modelling Tool
    The project team needed to establish the different types of equipment which use energy and how
    much energy (average) they used. In the past there was no real record held on the type of
    equipment in a specific room and the total amount of energy that room uses.

    The PTL audited all the rooms in C-Block counting the number of computers, printers, lights,
    distribution boards, network cabinets, electric heaters, air conditioning units,....etc. The audit
    included measuring the power consumed by all these items. Comparing this to the baseline data
    gathered from the sub-metering the PTL could account for 90% of the Kw usage throughout C-
    Block.



                                                  Page 14 of 37
Project Acronym: ECCILES (Energy Conservation in Computer Intensive Learning EnvironmentS)
Version:1
Contact: Derek Rout
Date: 30 April 2010



    Further to this all electrical items in C-block were shut down to see where the power was going on
    the site and to establish a true baseline. Portable meters were put around the hard-wired
    systems, and the data was logged. The exercise showed that even with all equipment off there
    was still a baseline consumption of energy. Looking at the history of the building it is a
    reasonable assumption to say that given the age of the building, there may be wires pulled
    through the building that even we are not aware of. Looking back 10-20 years wires for labs and
    classrooms were pulled to serve a power need, but not documented in the way it is done now.
    Tracing back the wires throughout the building is a huge task, but may not be a feasible use of
    resources at the moment.

    It is very easy to plug an electrical item in. When it comes to understanding how the energy is
    really used there are many gaps. The team needed to understand what is switched off
    automatically, what needs manual switching off or cannot be switched off for operational
    purposes.

    Building on this audit the team attempted to model the carbon emissions of the rooms throughout
    C-block. This was to find a tool which would further refine the data gathered and aid decision
    making when attempting future developments around campus, but also to identify energy
    expensive rooms and why. The process in a nutshell is described in the diagram below:




   The Energy Modelling Case study discusses further the data gathering issues with such a model.

   As energy monitoring can be refined down to the operations of a room given the correct data
   gathering. The data can further be detail the carbon cost per lesson, module, and even course or
   pathway showing a true carbon cost of delivering a course. This data if correctly held can support
   staff in finding ways of cutting back the carbon cost of running the courses. Before embarking on
   such huge data mining the case study highlights some of the resource difficulties in setting up
   such a tool and maintaining the data. A lot more work is required in creating such a tool and
   maintaining the data collection for the tool to act as a real time reflection of carbon consumption to
   activity ratios. If a university grows and the energy consumption grows, the tool can help establish
   if the carbon consumption per head is consistent or expanding too fast.

   This energy modelling tool can help staff analyse where the energy is most wasted and look for
   ways in which savings can be made. For example a lecturer uses a 40 seat lab to teach a class of
   20. Even if the computers are not on in the room it is still wasting energy. A room is at it‟s



                                                  Page 15 of 37
Project Acronym: ECCILES (Energy Conservation in Computer Intensive Learning EnvironmentS)
Version:1
Contact: Derek Rout
Date: 30 April 2010

   optimum at 90% capacity. Maybe the academic can pro-actively schedule a more appropriate
   room for the class. Building on that they can look at their optimum teaching group size before the
   quality of teaching is compromised and setup university labs which suit that capacity, limiting staff
   from having un-manageable class sizes.

         This modelling tool is a design, it still needs a lot more work, and research:
        into the data gathered to date further refined analysis to show periods of high and low
         demands, to optimise on services and minimise energy waste.
        understand other data requirements, what else do we need to be analysing?
        find a way of gathering that data automatically or very efficiently, reducing the need for extra
         human resources to maintain the tool.
        maintaining the data on a regular or ideally real time basis, the more real time the data the
         better the team can see the results from energy saving measures taken.
        keep and manage a changes log, as changes in activity can affect the raw data, again every
         piece of equipment in every room needs to be held on a data system and monitor them as
         they move around campus. Showing where the shift in energy has occurred. For example if
         someone moves from a single office to a communal office we can make savings by removing
         their printer and having them connect to the communal printer.
        establishing the data as an aid to strategic decision making across campus, strategically
         support campus moves to ensure that heat generating activities are in the cooler part of the
         campus and warm part of campus are better used in the winter reducing the need to over-
         heat or over-cool certain rooms.
        improve infrastructure by analysing the potential energy waste in a room before upgrading it.
        Devise a design for a smart campus, the data needs to support future improvements on
         campus. If funding is available to improve a block then there must be a host of suggested
         changes noted on the database which would improve the energy consumption for that room.
        Measure other activities such as energy consumed in remote learning. Make a comparison of
         energy use with on site teaching compared to remote teaching. Some would argue that
         students need the contact, how much contact do they need and how much can be done
         remotely, now with technologies such as Skype and video conferencing,...can a student
         conference call their lecturer and discuss issues, how will this affect the academic working
         patterns.

   As the project team dug deeper into the data it became clear that a lot of assumptions had to be
   made in deriving a carbon figure. The assumptions would have to remain until such a time when
   the data would be available to plug into the model. Initially the tool is aimed at helping staff and
   students find a way of improving the carbon footprint by making operational changes.

   The project team have learnt the benefits of gathering the data we have already got. The
   processes developed from establishing the base data have helped with understanding the way in
   which the courses are delivered. The University needs to further refine the data to cover all
   business activities not just teaching. In order to develop this tool any further would require more
   funding, resources and time.

5. Ventilation System Case Study
   Air conditioning is one the most energy hungry pieces of equipment in the rooms. The teaching
   rooms get very hot due to the equipment and the location within the building. There are two rooms
   which are fitted with air conditioning and are used the most. The other rooms are still hot but must
   be used as bespoke labs are setup in them. The heat generated over night really hits anyone who
   entered the room in the morning. A solution was required but the facilities group really did not
   want to add to the energy consumption in the area.

   Campus users complain about the conditions in the rooms and there is continual pressure to install
   air conditioning. The installation of air conditioning is a costly process and consumes a lot of
   energy. The Ecciles project is trying to reduce their carbon footprint, an alternative method of
   cooling needed to be investigated.




                                                  Page 16 of 37
Project Acronym: ECCILES (Energy Conservation in Computer Intensive Learning EnvironmentS)
Version:1
Contact: Derek Rout
Date: 30 April 2010

    The block has a mix of teaching rooms, staff offices, comms rooms and server rooms. There are
    many factors to consider when looking at the ventilation system. The facilities group were tasked
    with finding a free air solution to the Air Conditioning:

    The case study addresses investigation, feasibility testing and implementation.
    1. Engage Free Air consultant
    2. Establish Temperature monitoring across target rooms
    3. Design Free air solution
    4. Assess costs and feasibility of solution
    5. Install in test rooms
    6. Monitor effectiveness
    7. Produce effectiveness report

    The solution found was Evaporative Cooling. It works by passing the external air through a wet
    sponge in the system. The air passing through the sponge is naturally cooled by the cold water in
    the sponge. The water is refreshed periodically and the system automatically has shut down
    procedures which ensure that the water is kept clean. If the air needs to be cooler a fan passing
    the air through the sponge cooling the air further. This system can cool up to 10 degrees below
    the outside temperature, and give up to 15 air changes per hour in the rooms.

    The facilities team opted for additional sound dampeners to reduce sound disruption to classes.
    As a system it is not air conditioning, it brings the outside temperature into the room and the air
    flows out via windows. Air conditioning however rotates the same air around the room cooling it
    and pumping it back into the room. Comparatively the free air system only uses energy when the
    fan is used. If the outside temperature is sufficient to cool the rooms the fan is not needed and
    therefore no energy is used to move natural air through the building.

    The case study covers in detail the issues relating to installing such a system and the benefits of
    having the system in place.

    Installation
    The ventilation system is being installed in three phases. Installation will be disruptive to teaching
    and exams the majority of the installation works will be done in the summer 2010. Two units have
    been installed over the Easter break. One in the school admin office, and the other in a couple of
    teaching labs. The difference was noticed immediately by the students in the labs.

    The installation in the server room is the first free air ventilation system being installed in a server
    room across Europe. The benefits of the free air system are very big. The whole block can
    potentially be cooled with the same amount of power consumed by current air conditioning units air
    conditioning unit.
   Total energy consumption Evap Cooling                     996 kWh
   Total energy consumption Regular AC                    10358 kWh
   Total Energy saving from Evaporative cooling           9362 kWh
    The system can be run at 10-20% of the cost of running the air conditioning units.

    Installation is currently going on and the project team anticipate full installation of the free air
    cooling system by the end of the summer. This will be followed by monitoring and to produce
    effective reports. The reports can only be accurate after testing over a summer. The cooling
    system in class rooms will only be needed during summer months May-September. Over the
    winter the system will be pushing fresh air through the rooms but it will not need to cool the air.
    The cooling over winter will be monitored there may be instances of cooling needed depending on
    the heat generated by the computers in the rooms.




                                                  Page 17 of 37
Project Acronym: ECCILES (Energy Conservation in Computer Intensive Learning EnvironmentS)
Version:1
Contact: Derek Rout
Date: 30 April 2010


6. Awareness Case Study
    As a university Bolton has recently attempted to introduce green measures to the university
    infrastructure. Over the last 3 years the facilities department have attempted to make the campus
    more carbon friendly, slowly newer measures are introduced but it seems to be a general
    university perception that anything “green” is rolled out form the facilities department. As
    ECCILES has progressed it is clear that with all the departments working together there is a much
    bigger benefit to energy conservation. The team took a number of different approaches to raise
    the awareness of energy conservation across campus.

    The Awareness Case study goes into the whole approach in more detail. The team tried three
    different approaches to engage staff into adopting “greener working habits”.
          Bulletin Board Awareness – this included poster campaigns, bulletin messages,
             questionnaires... all designed to take a soft approach to prompting campus users to adopt
             greener habits.
          Forum Style Discussions and Debates – these include word of mouth, debates on a trip to
             a power station, discussions with staff around campus. Proved to be an effective tool in
             dispelling energy myths and gathering campus user perceptions. Also to convey project
             activities to staff. Most campus users have experience of having sat in a discussion and
             come away with ideas of how to make improvements. That was the idea behind the trip,
             encourage enthusiasm amongst those who came.
          Official Training Provision– This was a series of training sessions rolled out to different
             members of staff across all levels of the organisation to inform them of the carbon issues
             and areas where they can make real energy savings as individuals.

   All approaches were effective in their own way but not as effective as hoped. It is difficult to
   change peoples‟ habits. For example if the poster campaign worked as well as it did then all staff
   would be tuning lights off around campus. Unfortunately this is not the case. Members of the
   team found the forum style approach quite effective because you could really get into their
   perceptions and influence change in their enthusiasm and ideas. Of course there are those who
   will always be resistant but it is a great opportunity to break down their perceptions and reduce
   their barriers to change.

   A combination of poster campaigns and carbon training has benefitted the university in raising the
   awareness of how important carbon reduction is to the management of the university. As a
   university Bolton has a long way to go in changing old working habits. A soft approach is the initial
   step towards a cultural change in the university. The general awareness that this is important to
   the university as opposed to a select few trying to “go green” sets a precedence for all campus
   users to do their bit and adjust the way in which they work and approach energy.


7. WAKE on LAN Capability
    One of the perceptions from the academic staff members was that staff were not aware of their
    remote working facilities. Some professors connect to their workstation on their desks on
    campus. This enables them to access archived e-mails remotely. However this means that the
    workstation on site, have to stay on and outside of power management controls. As more and
    more professors do this the university will be right back to square one where everyone wants to
    remote connect to their PC‟s in the office therefore leave them on all the time. The university
    currently provide facilities to check e-mails via webmail and to access files and folders via online
    directories. If staff had better knowledge of this type of working they no longer have to make the
    journey onto campus to input results on the SITS system. However this would only be worth the
    university‟s efforts if staff rooms can be sectioned off and shut down for periods such as the
    summer and they use landing spaces if they need to come into the university.

    At present academic staff have not changed their working patterns enough for the university to
    start locating them in accordance to whether they would be on campus or not. Their offices
    cannot be isolated and shut down for holiday periods. However this is something the facilities
    department are looking at when departments are relocated on campus.




                                                  Page 18 of 37
Project Acronym: ECCILES (Energy Conservation in Computer Intensive Learning EnvironmentS)
Version:1
Contact: Derek Rout
Date: 30 April 2010

    Students have however taken remote connectivity up in their working patterns and regularly
    connect in to university workstations at night and work on programs through the night. Some
    students send their rendering jobs overnight, or queue to the render farm. Previously the
    university would have to had to leave workstations on overnight to allow students remote access.
    This added to our energy consumption figures. Since then University of Bath adopted a Wake on
    LAN system which allows staff or students to wake a PC up and connect to it. This is being
    tested in the computing labs within BCCT and the technician can wake the PC‟s up individually or
    in blocks to allow students to carry on working.
    Although some academics do not adopt this method of working it is growing amongst the student
    (the next generation of workers). It may even change the way in which a working day is defined
    in the future.


8. Survey of Lab Requirements and Student Demands
   Lab provisions at the University of Bolton are setup on a reactive basis rather than predictive.
   Technicians are changing software within two weeks of starting a semester and regularly
   thereafter. Some academics tend to want the newest and latest software to work with whereas
   others will use the same software until the upgrades dictate that older versions are unavailable and
   for students learning on two different versions it becomes difficult so it is easier to upgrade. All fine
   if there are not any software licensing and processing power implications. However, there is little
   or no predictive preparation work taking place to prepare for the semester in advance and given
   that the technician resource has reduced considerably the project team wanted to understand why
   such last minute decision making was taking place.

   Looking further into the problem the department has seen a number of changes with staff leaving
   and new skills being taught. Courses were split amongst existing staff and to adopt a module
   within weeks of it‟s start left academics stretched to ensure provisions are in place. The project
   team needed to investigate further into why software changes were happening mid-semester.
   Was it simply poor management or is the academic reacting to student demands? In order to
   encourage more joined up thinking the technician introduced a log for each module stating what IT
   provisions are required and when. This was to support the technician but to also get the academic
   staff to pro-actively think ahead for the semester.

   Understandably academics are focused on their professional disciplines and for them and
   converting this knowledge into the university business logic may be difficult if they have no
   experience of business disciplines. For example if a programmer is very good at programming
   and inventing software to perform complexed tasks, they may not have experience of applying cost
   benefit analyses, or seeing a real business benefit to introducing a desired piece of technology,
   such as a render cluster or new network facilities. Where an academic is looking to innovation an
   accountant would look to find the business gain. As a university the aim of this exercise was to
   strike a balance between the two perceptions, and not end up with “over-kill” on technology and
   energy consumption.

9. Powerman Case Study
    Despite awareness campaigns to encourage staff to turn off their PC‟s it wasn‟t making enough of
    a noticeable difference. Powerman is a piece of software which looks for keyboard and mouse
    movement. If this does not happen within a predetermined span of time the computer can be put
    into hibernate mode, and if the movement has not occurred for more than two hours for example
    then the computer can be shut down. These settings can be altered to suit an individual‟s
    working patterns. Initially Powerman was setup to:
          Monitor the system, see how staff were using the energy without any rules applied.
          Measure energy wasted through staff not turning off their computers.
          Collate data on activity and behaviour to establish the shut down rules “what is an
             appropriate amount of idle time before shut down?”

     Once all the setup and installation problems were addressed and the data being gathered was
     reliable, the hibernate and shut down settings were applied to every workstation on campus.
     Most staff did not realised that the system had been introduced. It made a big difference in


                                                  Page 19 of 37
Project Acronym: ECCILES (Energy Conservation in Computer Intensive Learning EnvironmentS)
Version:1
Contact: Derek Rout
Date: 30 April 2010

      places such as the library, 365 room and admin offices. In computer intensive environment it
      was more of a problem. Students in C-block have a number of different disciplines. Games
      computing, networking, ....etc. Networking labs are segregated from the university networks for
      student to build their own networks and test how their networks work ,the delivery of the taught
      material meant that introducing power management to the lab is difficult but is still a working
      progress. Games computing do a lot of rendering this makes it more difficult if powerman is
      always switching the computer off a more sophisticated piece of code is needed to look for
      processor activity as opposed to keyboard and mouse, the render farm software does deal with
      some power management and further testing is required to see if it fall in line with the university
      setup or if more code is required to sharpen energy saving abilities within the render farm.

      The deployment of power management across the campus can show conservatively an annual
      saving of £16,000. This can be greater with a little more fine tuning on the settings for shut
      down. Powerman does not currently shut down peripherals connected to the PC. The team are
      aware of this and have consulted with the developers of powerman to develop this capability.

      The Powerman case study explains these further.

      The university still needs to source a power management system for Apple Mac computers. As
      C-block did not have a Mac labs it was not an essential tool to acquire but remains an important
      part of the power management plan for the university.


10.      Virtualisation
      The university‟s server demands continue to grow and with relatively small workloads the servers
      still continue to (a) draw power, (b) need server room space, and (c) cooling. As a university
      Bolton needed to seek a solution to the growing server problem. Virtualisation is a solution
      whereby fewer high-spec virtual servers are purchased, and used to perform an number of tasks
      on one server, and therefore sharing resources between fewer „Virtual Machines‟.

       The university has been testing Virtual server since 2004, however as the technology has
      become more reliable and suitable for servers the more recent technology has been tested as
      servers. Due to the success of the testing the university has introduced three more virtual severs
      enabling the network team to replace the number of small workload servers for near-zero
      additional energy costs. The savings are evident in the space and cooling needs.

      So far the university has been able to reduce 65 servers onto 3 virtual servers and this figure is
      growing as the university continues to roll out more servers.

      Virtualisation has also enables the network teams to consolidate old servers. We have all faced
      the scenario where old projects have bought a server to carry out research/ project work
      specifically for that project, and once the project has ended nobody has de-commissioned that
      server. Virtualisation has enabled the team to do a “clean-up” task on old servers and work on
      newer maintenance policies of virtual servers and student/academic work. It has also affected
      the way in which the university approaches server procurement. This will be rolled out in time
      but is currently being discussed amongst those who use servers heavily. Having reduced the
      server space and energy usage it seems to have created a gap. It seems as savings are made
      they are seen as opportunities for more servers and work to fill that gap. It can be perceived as
      condensed capacity doing more work in a smaller energy efficient space, but then the challenge
      is to manage capacity at optimum and not allow the systems to reach saturation.
      Virtialisation Case Study.

11.      Server Room Temperature
                                                    0
      University‟s server rooms were kept at 15 C, and using a lot of energy to cool the room to that
      level. As the team attended various training events evidence showed that it was possible to
      increase the temperature of the server rooms, but was it as easy a task as just turning the
      thermostat up?




                                                  Page 20 of 37
Project Acronym: ECCILES (Energy Conservation in Computer Intensive Learning EnvironmentS)
Version:1
Contact: Derek Rout
Date: 30 April 2010

      The simple answer is NO. The server room management team needed to have the confidence in
      being able to increase the server room temperature. They needed to address a few matters first:

             How would changing the temperature affect the old servers in the room (temperature
              data provided is clearer on new servers but not old)?
             Bearing in mind that the university tended to look for the most cost efficient servers when
              purchasing, did we have any re-conditioned servers......Purchasing of servers is held
              centrally however the academics tend to source the suppliers, and spec for servers to
              suit what they are trying to do.
             How would raising the temperature affect the network teams‟ response time. A colder
              room will take longer to warm up therefore giving the network team the chance to get on
              site and follow shut-down / recovery procedures.
                   o This involved streamlining the recovery procedure
                   o Testing the rate at which the room would warm up given the heat emitted from
                       the servers.
                   o Did the server room need to be better arranged to allow optimum flow of cool
                       air?
             Which servers are essential to business operations and which are more vulnerable to
              temperature change.

      The vulnerability and sensitivity data of the servers needed to feed into the virtualisation strategy
      to ensure that the more sensitive servers were placed on a more robust, less temperature
      sensitive virtual server, enabling the team to eventually increase their confidence in the servers
      they manage.

      The server room temperature was increase a degree at a time. Sensitivity tests carried out at
      each increase in temperature. This was the preferred action plan to ensure data and vital
      services were not lost.


12.      Embedding
      As the team progressed through the TRIZ process, issues were raised outside of the project
      team‟s control. The university needed to have a holistic approach to resource control and retain
      the flexibility of subject areas to manage growth in their own schools. The team identified areas
      of improvement to ensure that the growth within computer intensive learning environments is
      managed and better controlled. By investing in the correct technology to improve the student
      experience and to effectively sustain energy conservation. The Ecciles team realised that having
      a central resource control would reduce the innovative flexibility of schools and maybe limit
      development, which can be debated to go against the university ethic of innovation. The
      possible task of managing the resources across the entire campus would require an extensive
      amount of data management and dedicated individuals to enforce the control. The university
      does not have such dedicated resources to spare and to have a central resource control may not
      support strategic growth of the university.

      The details of all the suggestions for management to consider are in the Embedding Case Study,
      the suggestions are listed as follows:
           Changes to procurement policy to invest in greener technology as opposed to the
              cheapest option.
           Carbon linked budgeting; influencing schools to operate greener habits by having
              rewards or penalties attached to the budget.
           Resource sharing across campus – encourage schools to share their labs to avoid
              duplication of labs on campus.
           Reduce active rooms during the summer or holiday periods. Have a smarter campus
              which would enable facilities to see the benefits of shutting down unused areas of the
              campus. For example instead of having 2-3 labs open with 2-3 users in each of them, to
              have the computers setup into one lab and close off the other labs. Saving on lighting
              and cooling costs.




                                                  Page 21 of 37
Project Acronym: ECCILES (Energy Conservation in Computer Intensive Learning EnvironmentS)
Version:1
Contact: Derek Rout
Date: 30 April 2010

             New Equipment justification – there should be a consultation process to see if by
              investing in new equipment. Addressing if it will affect the site‟s power loading, how will it
              impact on the sites‟ power backup systems, the full life cost of that equipment?......etc.
             Invest further into virtualisation. By not investing in many little servers from different
              schools or projects, why not pool the funds to invest in a virtual server. Or seek cloud
              space to meet processing needs.
             Introduce an in house induction programmes and then further invest in providing green
              modules on courses. Raising more awareness, and skilling students with carbon
              management skills they can take into their professional careers. By introducing this to
              the curriculum it demonstrates the university‟s commitment.
             Co-ordinated approach to IT management, setup labs to suit more than one discipline.
             Changes to the IT Policy; e.g migrate away from desktop printing and rely more central
              printing.
             Levy charges on new computer purchases or a central computer leasing system. To
              make campus users more aware of the cost of running the computers on site. E.g. in
              house render farm as opposed to the cloud rendering. What are the cost benefits? A
              better understanding of the running cost of aPC including its support and disposal cost.
             Green timetabling – by adding to the timetabling matrix with computer and software
              information, room efficiency information and seasonal efficiency information, the
              timetabling process can optimise on lab use for classes which need those specialist
              computers or software reducing the need for more labs because of availability, or
              optimise on the use of the most efficient rooms.

     The ideas is that one or two changes might make a difference, but a series of these ideas for
     example the training along with policy changes may improve attitudes to make carbon linked
     budgeting possible. A combination of all the changes would make a real impact on the energy
     consumption data.

     The team presented this information to the university management. With management support
     the suggestions can be discussed further, perceptions gathered and solutions refined.

A lot of different measures were implemented throughout the projects the technical solutions can be
measured for tangible cost savings and benefits. The operational changes will take quite some time
to implement, and will go through a number of cycles before a real implemented solution is realised,
most operational and policy changes require some cultural shifts and this is what takes the time.




                                                  Page 22 of 37
Project Acronym: ECCILES (Energy Conservation in Computer Intensive Learning EnvironmentS)
Version:1
Contact: Derek Rout
Date: 30 April 2010


Outputs and Results
The project has tried to influence energy conservation from a number of different angles in the
computer intensive environment. Some of the benefits have not yet fully been realised, it will take
some time for new systems to settle in and show some real benefits to the university.

For example with the lighting system took some time to be fully accepted as a different way of
operating. Some would say that it teaches bad habits and students would stop thinking about turning
lights off at home if everything on campus is automatically controlled. Others would argue that seeing
the lights go on and off will trigger a need to turn them off at home. Either way the auto lighting was a
must on campus to help the university improve overall power conservation. Current economical
climates are having an impact on future funding models where penalties will be enforced if energy
savings are not made around campus.

Each case study noted above presented their own energy saving results and are summarised as
follows:

    1. TRIZ – sometimes the end result derived from TRIZ may be an obvious answer, having gone
       through the process of TRIZ and come up with the conclusions as the team has done some
       could say that the solution could have been derived without using the processes of TRIZ.

         In this instance TRIZ has given the team a platform within which to debate the problems, filter
         out the false information or myths and find an optimum solution. As the cliché goes it is the
         journey not the destination which is more fruitful. TRIZ has served as an effective tool to
         strategically analyse the problem, and to strategically find a solution by predicting the impacts
         of changes implemented. The process by which this was done has lead to a better cross
         discipline sharing of information and working practises. Where one group could ask why it
         was not possible to do something, and another group highlighted the impact upon their day-
         to-day work or goals. This meant that a solution found was a collaborative solution which had
         a better chance of success. The TRIZ methodology forced interested groups to encounter the
         perceptions of others and challenge their own. It stimulated debate and encouraged a
         compromised if not totally agreed solution in almost all cases. It became a tool for people to
         interact better.

         Using this methodology IT, facilities and end user groups have been able to work together.
         Each group have a better appreciation of each others‟ areas of work. Having invested the
         time and effort into the TRIZ methodology the project team was able to come up with the
         operational solutions to the problem. When Professor Morton presented the same problem to
         Masters students in their TRIZ module, the students were able to understand the problem
         enough to come up with technical solutions but not operational solutions. For example Solar
         panelling, painting the roof white to reflect heat rather than absorb heat...etc. Operational
         changes have been realised as part of the solution through discussions in the TRIZ
         workshops. Ideas such as providing laptops to all students were simply infeasible because
         there is a greater security risks on campus, and that laptops wouldn‟t be able to provide
         workstation level services to meet student demands. This collaboration has resulted in a
         change in the way future projects are collaborated on.

         TRIZ takes a unique approach to defining the problem. For ECCILES this benefitted the
         group because it presented a methodical approach to defining the problem, and can focus the
         full team on a specific problem rather than the conversation going off on tangents. By
         separating the problem using the inventive principles the team were able to look at solving the
         problem in different ways. It has also been a great tool to get group “buy-in”. People are
         more inclined to be a part of the solution if they were involved in solving the problem.

         The use of TRIZ allowed the team to ask the questions which “fell on deaf ears” in the past,
         the application of the methodology driven by the funding criteria enabled the discussions to
         take place. Without which the university would still be at a point where they are feeling for the
         technical “low hanging fruit” solutions and being able to see the root of the energy
         consumption problem. By applying the methodology to the most energy problematic area on



                                                  Page 23 of 37
Project Acronym: ECCILES (Energy Conservation in Computer Intensive Learning EnvironmentS)
Version:1
Contact: Derek Rout
Date: 30 April 2010

         campus a wider spectrum of solutions was sought with the student experience in mind (the
         organisation‟s primary business focus).

    2. Metering – has become a very useful tool throughout the project. The base lining exercises
       have enabled the PTL to look at the contributors to the baseline and reduce the energy waste.
       With being able to meter and sub-meter the university can focus on specific areas to find and
       measure savings. The data gathered has been at the centre of being able to measure if
       energy savings are being made, and will help when producing data to funding streams.

         The sub-metering has been an in-valuable tool in simply understanding the energy demands
         of the equipment and the activities carried out in the rooms. When installing improvements
         such as lighting the sub-metering system was able to measure the savings made using the
         software which was developed with the metering system. The metering system can highlight
         sudden increases in energy use resulting form new equipment or even upgrades of
         equipment. The metering system can help when measuring overall energy consumption by
         schools, setting out carbon data to be presented to the schools who will then look for
         opportunities to make savings.

    3. Lighting – By changing the lighting in the block the project has calculated a 46% saving on
       cost of lighting.

                 Lighting             Before Lighting       After Lighting       Total Saving
                                          Change               Change
         Kw of electricity used       5474.25               2932                2542.25
         Kg of carbon                 3136.74               1680                1456.74
         Cost in £                    682.93                365.50              317.43

         These figures are based on the same amount of running time as the old lighting. Given that
         they are on sensors there is an even greater saving to be had. As mentioned above there are
         some recommendations to be drawn from installing the new lighting system. It is advisable to
         invest in the more sophisticated lighting sensors in order for it to be accepted by the staff and
         student using the area, and sustaining a suitable working environment. It was surprising how
         staff reacted to the removal of light switches, however with time the complaints have reduced
         and the system is accepted. Sometimes it requires a shift in working methods, for example
         adjusting to the way the new lighting system works as opposed to comparing it to what was
         there and finding arguments against the new system.

    4. Energy Modelling Tool – this exercise helped the project greatly in understanding and
       realising that there are a lot of environmental factors involved in energy conservation. There
       is still a lot of data mining to be done to develop a true measure of carbon for each room and
       breaking it down to activity. The model design has highlighted how rooms can be better used,
       and what activities are more carbon expensive. Ultimately the model could generate a carbon
       cost (be it per head, per hour, per class, module, course, pathway, school,...the list goes on)
       for the rooms, or even blocks, it could then be further refined to be used as indicators for
       carbon linked budgeting systems where savings are rewarded and increases in energy
       consumption have penalties. More discussions and perception gathering need to be done in
       order to establish how accepted such a measure could be. Most schools would argue that
       this would mean they cant grow because it would indicate an increase in energy consumption.
       The aim of the tool is for schools to find inventive ways of delivering courses and minimising
       the energy consumption. A key example of carbon measurement would be carbon
       consumption per student/ head. If the number of students increase and the per head ration is
       sustained then the school is maintaining a level energy consumption, however the inverse is
       also possible where if the number of students drop for a course, then has the energy
       consumption dropped?

         With HEFCE dictating funding reductions have to be made and that is absolute goals, school
         will have to aim for total absolute reductions overall to meet HEFCE funding requirements.




                                                  Page 24 of 37
Project Acronym: ECCILES (Energy Conservation in Computer Intensive Learning EnvironmentS)
Version:1
Contact: Derek Rout
Date: 30 April 2010

         The process of performing the audit and shutting down the systems to try and get the
         consumption down to zero has been a great benefit to the team in understanding where the
         energy went in the building. Surprisingly enough the complete shutdown of C-Block was a
         more difficult task, because after turning everything off and accounting for hardwired systems
         the technicians still could not get the energy consumption down to zero. This could be due to
         old wires in different parts of the building, or there may have been some under the desk PC‟s,
         or chargers,...etc missed in the shutdown. It highlights the extent to which energy consuming
         technology is unrealised. How often have you had something plugged in and not realised it in
         the corner? Will all campus users be able to know exactly what is plugged in their work
         space? Do we culturally take the supply of energy on demand for granted?
         The asset accumulation over many years has made the task of shutting everything down very
         difficult. As a result of this exercise the PTL has introduced assets, their upgrades, log simply
         to help him understand the changes to the baseline and to support future energy conservation
         strategies.

    5.    Ventilation system – Instead of Air conditioning the facilities team have sourced evaporative
         cooling to be installed in the buildings. This system is still being installed, and is the first in
         Europe to be put into a server room. The system will be installed over the summer, and be
         fully functional by October. The project team will then measure and plot the power
         consumption of the system for a whole year to get some real tangible data reflecting on how
         effective the system is over the summer periods. October will be the start of autumn,
         meaning less need for the cooling element of the system more the flow of air through the
         rooms. Peak summer is from May to September.

         The estimated benefits of the system are :

                 Total energy consumption Evap Cooling        1972 kWh
                 Total energy consumption Regular AC         10358 kWh Based on 10.20% cost of AC
                 Total Energy saving from Evaporative cooling 8366 kWh

         The system can be run at an estimate of 10.20% of the energy used in running the air
         conditioning units. Actual data will be presented once the system is in place. The project
         aims to continue to report this data via the project website.

    6. Awareness Raising – Prior to the awareness campaign energy conservation and carbon
       reduction was an “option” for staff in the university. Recycling and turning lights off was
       “something you did if you wanted”, those who were carbon conscious did take measures to
       turn off equipment and recycle but they were a minority. The awareness campaigns although
       they raised the awareness of energy saving and made campus users aware that it was an
       important issues to the management of the university most felt it was still the responsibility of
       someone else.

         The poster campaigns encouraged campus users to turn things off or even change the way in
         which they work, the campaigns raised the importance of energy conservation around the
         campus. It portrayed a message that energy conservation is important to the management of
         the university. The momentum of the training programmes delivered needs to be taken
         further into in house induction programmes. The team aims to recommend developing a
         series of induction programmes and green teaching modules which are targeted at using
         some tools from TRIZ, to come up with operational solutions, find greener ways of working
         within the university. Coming away from always seeking technical solutions, and by using the
         technology we currently have in a smarter way, changing processes and procedures to
         become more carbon prudent and by engaging campus users in finding a solution may
         improve the success of implementing further carbon reduction measures. Most importantly
         reducing resistance to energy saving measures introduced on campus, and ultimately
         reducing the risk of failure to adopt energy saving measures.

         When team members engaged in energy saving discussions with other campus users, the
         topic is more widely discussed around campus. When in the past the discussions about
         policy change were not possible are now opening up debate about how feasible certain



                                                  Page 25 of 37
Project Acronym: ECCILES (Energy Conservation in Computer Intensive Learning EnvironmentS)
Version:1
Contact: Derek Rout
Date: 30 April 2010

         changes around campus could affect the way an individual works or perceives the problem/
         possible solution. The discussions can now take place.

    7. Wake On LAN – As we grow into a 24 by 7 connected society it has become increasingly
       popular for staff and students to access their workspaces at any time. By using Wake on LAN
       the university can reduce the need to have computers on all the time. This is still in it‟s testing
       stage and will impact on the way almost 150 plus students access the university workstations
       out of hours. Once the testing is done and can be rolled out, it will be added to the induction
       programme as another way of accessing work outside of university operational hours.
       However in the meantime the project group needed to address the long opening hours of labs
       and them only being used by a select few in each room.

         There was a clear technical and security need to reduce the lab opening time back down to
         5pm closure as opposed to the current 9pm plus closure. The fact that the labs were being
         cooled and lit for the 4 hours extra every day of the semester needed to be reduced. As
         explained above the idea was to introduce lab spec workstations into the 365 room which is
         open 24/ 7. By redirecting student to these workstations the lab technician could do an
         appropriate shut down of the labs at 5pm before leaving the site, saving the university on
         energy costs and security risks. It is recognised that there are certain times in the semester
         when workstation demands will be high and labs will be opened and available to students
         then. The 365 room coupled with the remote connectivity enables the students have 24/7
         access to university services / workstations.

    8. Survey of IT Lab requirements – As lab setups were changing very regularly within a
       semester it was important to encourage IT academic staff to start thinking ahead in terms of
       resources. This would encourage staff to source the best software and resources in good
       time and not at last minute. Also to develop a lab sharing strategy to come away from
       bespoke labs to labs which can be used by a number of disciplines. A couple of forms were
       introduced to better inform the technicians of the lab setups required and to help other
       departmental users to access workstations and software facilities. Technicians could then
       setup labs to strategically support more than one subject areas/ pathways. The aim was to
       encourage academic staff to start forward thinking in their teaching semester and to share
       their thinking behind the setup. The forms are in their early stage of implementation, some
       staff found them a good idea and others did not. The information given could help the Deane
       of the school better share facilities and strategically resource the school better for future
       courses. The cross campus sharing is one of the recommendations which will be put forward
       to the management of the university to implement.

    9. Powerman – A simple solution to implement however once it was set with the shut down
       routines it was quite surprising how many members of staff actually remote connect to their
       own computers as opposed to on-line access. Powerman with conservative calculations can
       save the university £16000 within a year. This is an incredible amount of power being saved
       by simply automating hibernate and shutdown procedures in the university. There is still
       scope for further savings to be made by fine tuning the shut down settings.

    10. Virtualisation – by migrating 60 servers onto one big server the team have seen a
        considerable saving in both power and server room space. Old technology did not have
        energy conservation built in so to replace them with an efficient server has made a big
        improvement on the amount of power being consumed. It‟s also reduced the strain on the
        UPS system. If all the servers were virtualised then this could potentially save the university
        approximately £12000 per year. It would also enable the university‟s networking team to
        introduce consolidation strategies for work held on old servers. A clean up process in a
        nutshell, presenting opportunities for the university to possibly adopt the use of external cloud
        technologies, for general tasks such as hosting student e-mail and workspace services. Or
        possibly move the staff e-mail service onto the cloud, again there are some possible issues
        with this which need to be investigated further.




                                                  Page 26 of 37
Project Acronym: ECCILES (Energy Conservation in Computer Intensive Learning EnvironmentS)
Version:1
Contact: Derek Rout
Date: 30 April 2010

    11. Raising the server Room temperature – the team wanted to see what would get worse if
        the server room temperature was raised, also it was a way of testing the server room to see
        how it would behave with “Evaporative cooling”.
              did the temperature have to remain at a specific level or could it fluctuate?
              How would the temperature change affect the response times for the server room.
              How will the older servers react to the warmer environment and would they be the
                 most volatile in the event of a cooling failure.
              The server room reached a confident temperature of 22 C. The response time still
                                                                           o

                 needed to be balanced with the backup battery support to ensure there is sufficient
                 time to do correct shut down procedures on all servers before complete power down
                 or air cooling failure. Once the cooling system is in the temperature sensitivity of the
                 servers will need to be tested.

    12. Embedding – The embedding information still needs to be presented to higher level
        management, unfortunately it is a suggestion opportunity and although a combination of
        solutions would show real results the management may or may not choose to take some of
        the suggestions on board.

         Some suggestions are being implemented in the university:

                  Green Procurement plan – the procurement officer in the university is in the process
                   of developing a green procurement policy which will incorporate travel, energy
                   consumption and procurement of new technologies.
                  Co-ordinated approach to IT Management - The IT policy is being re-written to
                   incorporate energy saving strategies within the organisation. For example phasing
                   out desktop printers and linking all staff to shared printers or photocopiers. Moving
                   away from academic staff having a laptop and a PC, introducing the FIT PC‟s into
                   areas where it suits the purpose reducing the energy consumption for simple
                   purposes such as the OPEK computers in the library....etc.
                  Virtualisation – The virtualisation strategy is dependent on available funding and as
                   the funding becomes available it is invested in this strategy. The need for further
                   virtual investment has been highlighted to the management of the university.
                  Sharing of Resources across campus – The work done in the ECCILES TRIZ
                   sessions has set a ball rolling where the technicians from different schools are able
                   to discuss the topic of resource sharing, and maybe have a shared resource in the
                   365 room. At the moment this is an agreement between technicians and the
                   decision needs to be firmed up within the schools and make it a school strategy, this
                   will reduce the intervention of staff who may have a territorial claim on the labs.
                   There are potential issues with this system including which school pays for what
                   equipment and if a common resource sharing could be really be achieved amongst
                   academic staff.

         There are still some more embedding strategies which if implemented would help the
         university achieve its energy saving goals, and move further into strategic development of the
         university.

                 Induction programme for all staff and students – The idea here is to demonstrate
                  practical changes campus users would have to make in order to go green. Giving
                  them techniques that they may not have know about for example printing
                  confidentially from the photocopier, develop an office shut down policy, making the
                  small changes to working habits which will result in energy savings. Showing campus
                  users that this is the way to work in the university, empowering staff and students to
                  change the way in which they approach energy as a resource.
                 Following on from the induction programme the university can then start to tackle the
                  organisational changes to saving energy, gathering more data on energy modelling to
                  feed into carbon linked budgeting strategies. Hopefully then will staff be able to look
                  further into finding opportunities to reduce energy waste. Leading onto green
                  timetabling, capitalising on opportunities to shut down areas of the campus during
                  vacation times, writing in energy, equipment, carbon cost which developing courses


                                                  Page 27 of 37
Project Acronym: ECCILES (Energy Conservation in Computer Intensive Learning EnvironmentS)
Version:1
Contact: Derek Rout
Date: 30 April 2010

                  in the future, and further still embedding green techniques in the professional skills
                  the academic teach, meaning student are then equipped to apply green techniques
                  out in the workplace.

         20% Energy Saving Made in the Pilot are –

         So far the real measurable savings made by the project are:

                 Activity             Saving in Kw          Saving in Kg of            Saving in £
                                                               Carbon
           Lighting                2542.25                 1456.74                 317.43
           Ventilation             9362
           Powerman                                        1600                    13333
           Virtualisation          144                                             12000
           Total

The project has made measurable and immeasurable savings in energy consumption within computer
intensive environments. Ecciles has also realised that the computing industry is driving up the energy
consumption by increasing processing power requirements of, software and platforms which require
                                      rd
more rendering power. Digital and 3 generation capabilities have increased the demand for better
processing power, increasing the reliance on energy. As a university academic staff must access up-
to-date technologies in order to keep up with the industry and teach with the platforms which are used
out in industry. This creates an increasing demand on the university‟s technology portfolio and how
much energy they require. Comparing a lab workstation today to one 3-5 years ago, a newer PC may
have energy save modes built in but when they are used they draw more power compared to their
predecessors. As the university continues to grow so does campus facilities. These facilities are
anticipated to grow and the university‟s strategic growth needs to be managed in order to keep energy
consumption and carbon emissions as low as possible.

The main problem faced by the computer intensive environments is growth. When the initial
measurements were made there were approximately 300 students using that workspace, now the
school has increased its student number by 50% and labs have been more heavily used and
timetabled to meet the teaching demands. As the industry expands so does the processing
requirements of the equipment leading to more power being consumed, and fewer shut down
opportunities for the labs. Care must be taken as to balance optimum use of on-site facilities without
compromising the student experience, ensuring that the lab use / demands do not exceed saturation.

Outcomes
The project team set out to make a 20% saving on the energy consumption within C-Block. At the
time it seemed to be an easy task because the university could identify the “low hanging fruits” for
saving energy. The measures adopted were ventilation, metering, lighting, energy modelling, Wake
on LAN, Surveying lab requirements, powerman, virtualisation, raising the temperature in the server
room, awareness raising, introducing machines in the 365 room for out of hours access, embedding
solutions into university policies such as procurement policies, IT policies, green induction training,
green modules, carbon linked budgeting,

Referring back to the project plan the critical success factors were measured against the following:
    1. A detailed model exists for the energy consumption factors around computer use at Deane
         Work done with metering, powerman and energy modelling all contributed towards
         establishing energy consumption factors around C-Block. All the factors and data collected
         gave the team a better insight into how energy was consumed in the computer intensive
         environment. The team went about analysing the data to find where the waste occurred and
         to try and solve the problem. It was surprising to see how the energy was being used in the
         labs, and obvious solutions were introduced to tackle the energy waste.
         A combination of meter baselining data, room audit data and data from power-down exercises
         fed into an energy modelling tool. The figures build a better picture of C-block‟s energy use.
         The block can now be understood better as an entity because:


                                                  Page 28 of 37
Project Acronym: ECCILES (Energy Conservation in Computer Intensive Learning EnvironmentS)
Version:1
Contact: Derek Rout
Date: 30 April 2010

                 Energy modelling data established the ability to find the carbon cost of the rooms and
                  then compare the data with other rooms.
                       o   This can be further refined into:
                       o   Carbon cost per person data
                       o   Carbon cost per class
                       o   Carbon cost per module / course/ /pathway
                       o   Provide comparative data for onsite teaching vs remote
                       o   Streamlining site use to ensure staff use rooms at their optimum
                       o   Rating areas of the campus in comparison to other “at a more refined level”
                       o   Rating departments and their ability to cut carbon emissions
                 The data can be aimed at optimising use and minimising energy waste. There are a
                  lot of thing that can be done with the correct refinement of the data. The tool is not
                  fully developed, the Energy modelling case study details the experiences of the team
                  when attempting this tool and some areas which need further funding to create a
                  feasible and transposable tool for the whole community.
    2. A 20% reduction on the energy consumption due to computer use at Deane in its current
       configuration.
         The project aimed to take power usage reading at the beginning and then at the end of the
         project. With all the measures taken by the project to reduce energy consumption
         (powerman, virtualisation, increasing server room temperature, ventilation,...etc). Each
         change made was then monitored to establish the energy savings made.
         The project to date has made the following savings in energy:
         However there are some things still not fully implemented. Ventilation, Power management
         on MAC labs, implementation of some of the embedding strategies,...etc.
         The ventilation system and MAC powerman still need to be fully installed before savings are
         realised. Until the ventilation system is in place and the actual data is gathered the team
         cannot evidence real savings made and the impact the energy saving measures have had on
         C-Block.
         There is still a danger of no matter how much of a saving is made, there will always be a
         growth factor counter acting against real energy saved. This does not mean that the team
         should be disheartened because having made the savings additional activities are possible, if
         the savings were not made the energy consumption would most probably be a lot higher.
         There has been an increase in the number of students using the area this year compared to
         last year, computers have been given upgrades, a new in house render farm has been
         developed, labs have become more specialist rather than universal in their setup, more
         students access the lab machines remotely this year than in the past. More academic
         subjects have come on board, and more non-computing disciplines are now digitally driven.
         A costed plan is in place for optimising the temperature in computer teaching rooms at Deane
         The project aimed to solve the overheating problem in C-block. Facilities did not want to
         simply install more air conditioning units because they are very power hungry, the team
         wanted to find a free air solution.
         Evaporative cooling was the feasible solution found, with a reasonable payback period.
         Installation is due to be completed by late October 2010. Any real cost benefit data can only
         be analysed once the system is in place and monitored. The system works by pushing fresh
         air through the rooms which health wise a much better solution to air conditioning. The free
                                                     o
         air solution will cool the rooms as low as 8 C below the outside temperature and has the
         added benefit of supplying 12 air changes per hour, which is not supplied by AC alone. The
         feedback from students and staff has been very good. The conditions in the rooms have




                                                  Page 29 of 37
Project Acronym: ECCILES (Energy Conservation in Computer Intensive Learning EnvironmentS)
Version:1
Contact: Derek Rout
Date: 30 April 2010

         improved so much that they are being used for more activities e.g. exams for special needs
         students, which was unheard of in the past.


    3.   Staff and Student have greater awareness of the issues of energy consumption and there is
         evidence of changed behaviours
         The awareness campaigns raised energy conservation around the campus. Academic staff
         are more formally encouraged to look for greener options when investing in technology.
         Technicians to embed power management into the networks they manage and introduce for
         example rolling out a render farm with power-down capability when not in use, this took longer
         to roll out but understandably important to implement as a benchmark code of practice for the
         department.
         Campus users who came on the trip to Electric Mountain were keen to get involved in a green
         forum where discussions about potential campus changes could be hosted. The team are still
         looking into setting this up for the university, ownership, and administration need to be
         managed on a daily basis. This will be a part of the project‟s exit strategy.
         Proactive behavioural changes have been very slow. However the different soft approaches
         used have raised the profile of what the ECCILES team are trying to do, the importance of
         green issues across campus, and the perception of it‟s importance to the university
         management. The carbon trust training was expected to have shifted management behaviour
         to change some internal school policies, but this is quite slow in progressing. However what it
         has done is open the discussion of change, warming campus users to the different changes
         which can occur and why they are necessary to the university.
         The Embedding strategies are still to be rolled out. If successful in getting higher level
         management support it may be that behavioural change will occur more quickly. Some of the
         suggestions such as policy changes are currently being implemented but have not yet been
         rolled out. Initially it is recommended that a university wide induction programme is adopted
         to show campus users how to actually be green and basically “this is the way we adopt a
         greener way of working at the University of Bolton”. The underlying reason behind such a
         programme is to gather perceptions but to also address the changes which will occur around
         campus. Most campus users can realise the benefit of changing lights and ventilation, they
         cannot see the energy benefits of organisational change, strategic change and changes to the
         they work. Following that strategies, such as carbon linked budgeting or green timetabling
         may be easier to implement.
The project has achieved a lot of different outcomes, the successes are:
        The application of the TRIZ methodology to the problem succeeded in bringing together a
         multidiscipline team to establish collaborative solutions to the problem of energy conservation.
         The team were able to thrash out the problems and eliminate barriers to finding a solution.
         Through this methodology, finding alternative ways of solving the problem, improving the
         success of solution systems and most importantly reduce resistance to implementation.
         TRIZ was a time intensive process and required a considerable amount of commitment from
         all team members. This was difficult when trying to get all members together at the same
         time. Academic involvement reduced as the meeting moved on however there was a strong
         investment from Facilities, IT staff and the School Technician who is in charge of the labs.
         If the TRIZ process was to be done again it would be more streamlined and focussed on the
         subject to hand. The structure allowed for the teams to go away and study or produce results
         for the next session. The consistency was important as to not having to repeat discussions
         for those who missed the sessions. It is difficult having to repeat the discussion from the
         week before.
        Rollout of powerman, virtualisation, ventilation, lighting and metering were all aimed at
         improving the energy conservation of C-Block. Each of which have had their own technical
         learning experiences with installation and how campus users respond to the systems. The
         case studies detail the way in which the systems were implemented.




                                                  Page 30 of 37
Project Acronym: ECCILES (Energy Conservation in Computer Intensive Learning EnvironmentS)
Version:1
Contact: Derek Rout
Date: 30 April 2010

        A combination of sub-metering and circuit audits has established the baseline power
         consumption. By looking at the equipment plugged in and when they are used, the team
         could look for ways to reduce the baseline energy consumption. For example reducing
         weekend consumption of energy by ensuring that unnecessary equipment is switched off.
         The perfect example is the photocopier. Introducing shut down procedures for each area on
         campus would reduce the carbon wasted over the weekend.
        Energy Modelling was designed (but does need a lot more work) to be an effective tool in
         measuring the energy saving efforts of departments, courses, modules...etc. By
         benchmarking in one year using metering data to find energy saved. By refining the data
         gathering further, as new ways of saving energy are realised. The tool can eventually lead to
         carbon rating courses and pathways for cross comparison. The benchmarking can also be a
         useful measure for carbon linked budgeting to encourage school to control their own carbon
         footprint.
        Research communities have probably done policy changes in some form or other, JISC have
         sites such as procureweb, which help with green procurement policies. The procurement
         policy within the university would be a control system, coupled with training would encourage
         campus users to apply energy conservation when making equipment selections, and justify
         why they would invest in that particular equipment. Consult with key departments before
         purchase to ensure it can be run on site, and state the length of time the equipment and data
         should be maintained by the networking teams. Every piece of computing equipment have
         also a need for backup and power backup.
        Introduction of equipment planning when creating/ validating a course. Using course
         requirement forms to set out equipment and software requirement schedules for the
         semester. This will reduce last minute lab setups and iterations of lab changes throughout
         the semester, encouraging academics to think a year ahead for their courses. Removing the
         reactive resourcing which, currently occurs. This advanced planning must include
         investigating the industry and software houses to see if they are planning on changing their
         material which would be available to the students.
        The project is built up of a lot of case studies, each case study shares the Ecciles team‟s
         experiences when implementing solution systems. The case studies are designed to show
         how Ecciles has addressed the problem, the problems the Ecciles team have come across
         and how the project team have dealt or not dealt with it.
        It is important to highlight that some of these measures in isolation can make a small
         improvement to reducing energy consumption, but Ecciles aimed to have sustainable usable
         systems in place, therefore the process of gathering the audit data, has also been a clean-up
         process of old systems which have never been de-commissioned or come off line and
         essentially unplugged.
        The areas still in progress: Ventilation is to be installed over the summer across both floors.
         FitPC‟s and thin Client solutions are still being tested, Wake on LAN is being tested,
         Virtualisation is subject to university‟s available funds, Embedding ideas are subject to
         approval and support from management and Surveying lab requirements is an on-going
         process, which will be refined with time.

Experiences – ECCILES has been a developmental learning curve for the university. The university
has been able to use the TRIZ methodology to understand the carbon issues better within the
university. Coming from not being able to discuss or implement change across campus to being able
to have discussions about blanket policy and strategic changes to incorporate energy as a resource
into the university‟s systems is a big improvement.

The TRIZ approach enabled the disciplines to gather and implement changes but it also enabled a
better involvement of the different disciplines to the problem. It highlighted some of the bad decisions
made in the past, which would not have come to light without the collaborative approach. The
approach has reduced barriers and has created a more collaborative approach to university activities
in the future. It is useful when anyone is struggling to implement a solution developed. By using the
TRIZ approach with the implementation team/ end users might highlight the flaws in either the solution
or the current way of working. Breaking down any false perceptions and addressing the real issues


                                                  Page 31 of 37
Project Acronym: ECCILES (Energy Conservation in Computer Intensive Learning EnvironmentS)
Version:1
Contact: Derek Rout
Date: 30 April 2010

as to what is stopping the problem being solved and what is getting worse by solving the problem.
The multi-directional approach tries to reduce the risk of making something worse which you did not
realise.

The aims of the university moving forward, are to develop some of the skills learned from the
ECCILES project such as feasibility testing for green solutions, business analyses of organisational
behaviour, energy monitoring for resourcing purposes,....etc and to develop them into the university‟s
teaching portfolio. Teach greener building designs, or greener network architecture, green data
management strategies, greener server room configurations, and possibly encourage greener coding.
The university realises it‟s responsibility to providing professional training and wants to be able to
teach these skills in all subject areas. The university students will be better equipped as the
professionals of the future to be able to take these skills into employment or their own businesses and
implement these practices. As IT skills are now the norm in industry so too will be carbon saving
skills.

Most organisations in the community have already implemented measures such as lighting,
virtualisation, power management,....etc. The ECCILES project has been able to improve a computer
intensive learning environment and can develop new strategies, for example to set optimum class
sizes, implement different server management strategies, improve the architecture within server
rooms, network and backup the equipment better, have a better realisation of on-site saturation,
changes to IT policies, changes to IT procurement strategies, implement subject linked access to
limited resources in open access computer spaces (365 room), find green timetabling strategies to
maximise on the carbon friendly areas of campus, and to improve the carbon footprint of old areas of
the campus.....the strategies are endless, all in order to save energy and use campus space at it‟s
optimum. Not all the ideas can be implemented immediately but they can be included when a change
opportunity arises. When a room is being renovated or when a department is being re-located.




                                                  Page 32 of 37
Project Acronym: ECCILES (Energy Conservation in Computer Intensive Learning EnvironmentS)
Version:1
Contact: Derek Rout
Date: 30 April 2010


Conclusions
ECCILES aimed to deal with energy conservation in an area which is influenced by an energy hungry
industry. As the IT industry is growing in processing power it is hard to restrain the growth of
computer intensive learning environments without compromising the quality of the subject being
taught. Students would prefer to learn the up-to-date technologies enhancing their chances for
employability and entrepreneurship.

At the start of the project the team had no idea as to how much energy was being consumed by the
Computer intensive environment or the factors which affect the energy being consumed throughout
the year. With the sub-metering system the project team could analyse the energy used and identify
energy spikes and map them to activity. By investigating further into the energy spikes, the team
could analyse that use and determine the level of necessity for that energy use. For example a server
compared to a printer. Which is more important and which can be switch off on evenings and
weekends. The perceptions gathered around that decision aided a better understanding as to where
the energy is used and why. The ECCILES project as a whole has improved the university‟s
understanding of energy being consumed around the campus, but also why the energy is being
consumed the way it is? Better solutions were found once the team started to ask why and if by
changing things are we improving the services, or are we creating more harm.

The ECCILES project used TRIZ to understand the energy problem of computer intensive
environments and come up with feasible affordable solutions to the problem. By using TRIZ the
project team addressed some of the internal barriers when trying to solve the problem. E.g.
departmental collaboration, end user buy-in, multi-disciplined approach to looking at the problem and
finding the optimum solutions to the problem. The use of perceptions was particularly helpful because
it helped break a problem down into:
       why it cannot be solved and
       by solving the problem is it creating another problem elsewhere?

Those seeking energy related solutions in the past were faced with contradictory agenda and could
not achieve a reduction or control of resources allowing control of energy wastage. Labs were
increasing due to configuration clashes and energy was not a part of the decision making process.

One of the reasons why TRIZ worked for ECCILES was because of the multi-disciplined approach,
taking all the arguments into consideration to find a solution. Initially the thinking was around
physical reduction of equipment/ labs or even shut down strategies for the physical equipment.
Towards the end of the project the thinking was geared around business operational and
organisational changes to support energy conservation. The geared thinking around TRIZ aided
more refined discussions around specific perceptions and develop solutions which would best suit the
organisation.

The solutions were at three different levels: Technical, organisational and operational.
    The technical solutions were pretty easy to implement, by introducing hardware or software to
        reduce energy waste.
    Operational solutions required a considerable more effort in discussing with campus users as
        to why the solution is designed the way it is, and implement a change in university culture. By
        gathering perceptions or opinions as to how feasible a solution could be, and analysing if the
        benefits out-weigh the harms created by solving that problem ( a risk analysis in effect, but
        this also included opinions/ perceptions which seem to have a strong impact on if a solution
        succeeds or not).
    Finally organisational solutions by changing strategies and policies of the organisation, which
        required higher level buy-in to the problem and solutions. Policy changes would impact on
        procedural changes to include energy at the decision making level of the university.
        Incorporating carbon/ energy, as another resource which needs to be managed at decision
        making. This can affect the way in which the university delivers it‟s degrees, should the
        campus radically change the computer labs, should the university migrate everything to the
        cloud? Should the teaching model stay the way it is? On-site vs remote teaching.....etc. All
        debates which need to be undertaken by the academic schools, it is stressed that carbon and
        energy savings have to be made.



                                                  Page 33 of 37
Project Acronym: ECCILES (Energy Conservation in Computer Intensive Learning EnvironmentS)
Version:1
Contact: Derek Rout
Date: 30 April 2010

As a university it is important to ensure sustainability, and future funding will be linked to the
university‟s ability to save Energy/ carbon. Some funding groups have already linked funding to
carbon and others are setting audit requirements which include the carbon cost of travel and delivery.
Funders are becoming more concerned about carbon footprints and the state commitment to reducing
carbon emissions. All involving elements of data which needs to be gathered and analysed to see if
the university is achieving the state set targets and also the market set standards (students recognise
that the next generation of businesses are going to be green orientated in the way they work, and the
ethics they follow in the way they carry out their work. For example in the past an employee could
have left their PC on all the time regardless of the energy consumed for that day. Now staff will have
to recognise periods in their work patterns which would allow for shut down).

This project has enables the team to recognise data gathering issues and how in the future that data
will be captured. Throughout the ECCILES project the team have been able to explore methods of
data capture and reporting, which will become a vital part of future reporting. If ECCILES as a project
did not happen the university would have been on a steep learning curve with the way in which carbon
is reported. ECCILES has given the university the ability to focus analysing the green information the
university needs and how that data is read and reported and acted upon.




                                                  Page 34 of 37
Project Acronym: ECCILES (Energy Conservation in Computer Intensive Learning EnvironmentS)
Version:1
Contact: Derek Rout
Date: 30 April 2010


Implications
The project has developed a strong exit strategy, aiming to implement a lot of the techniques
suggested in the Embedding case study. Not only should organisations look at the physical aspects
of saving energy and changing the culture within the university to reduce energy consumption but to
also looks at the way in which they “do business”. The ECCILES project set out to solve the problem
on the physical consumption of energy, trying to identify the waste in the physical rooms and set
measures in place to reduce that waste. When digging further into why the setup was the way it was?
To purpose that energy was being? Why were the computer environments growing so much, why
they were being changed/ upgraded and improved so often?

Looking at the problem on a holistic level there were other influences which affected the way in which
the computer intensive environments changed. Factors found were the industry, student preferences,
staff needing to keep up with the developments in industry and the way it all fits in with the teaching
model of a university teaching format.

The Embedding Case study highlights the strategies suggested which were outside of the project
groups control and needed to be considered by the higher level management of the university. The
suggestions were:
     Green Procurement policies – encourage more greener buying rather than going for the
        cheapest option.
     Carbon linked budgeting – Encouraging schools/ departments to find their own ways of saving
        energy by linking their school budget to carbon measures. If the school saves on energy the
        better their budget looks.
     Resource sharing across campus – getting departments to make their 60%ish used labs to
        allow other departments to schedule use of those labs.
     Course development to include resource requirements – realistic resource requirements are
        addressed at validation, enabling the university to accommodate the course within the
        resource they have or see a justified need for extra resources, labs,....etc. A streamlined
        approach to fill gaps in resource use. Optimising lab usage and availability.
     Reduce active rooms during vacations – Being able to shut down two labs and consolidate
        into one when demand is low.
     New equipment justification – Large investments such as powerful equipment or farms or labs
        need to go through a consultation process where the power support, backup support, housing
        support are all addressed.
     Virtualisation – investing collaboratively amongst schools and project to get virtual servers as
        opposed to many small bespoke servers.
     Awareness raising – Developing an in house induction programme which highlights the
        energy measure put in place on campus and to develop a cultural ethic for campus users to
        save energy.
     Co-ordinated approach to IT management – It policies to incorporate green measures, opting
        to spec a greener facility for departments and schools before “going for the highest spec”
        How many PC‟s are really used to their maximum specification?
     Levy or leasing charge on computers – to encourage schools and departments to consider
        the full life cycle of purchasing equipment. Encourage project equipment to see the cost of
        supporting that equipment beyond the life of the project. To make campus users realise the
        support and disposal cost of owning that PC, server,....etc.
     Green timetabling – using carbon data to allocate rooms, making sure the more efficient
        spaces are used and to their optimum capacity.
    
These are all techniques the team have suggested to the management of the university to further
develop and implement in order to see a greater improvement in the way the university saves energy.

        The project team started to develop an Energy Modelling tool which further refined energy
         use in specific areas. If the data was collated across the whole campus, the model would be
         able to map the use of space by a course module or pathway to realise the real carbon cost of
         delivering courses.




                                                  Page 35 of 37
Project Acronym: ECCILES (Energy Conservation in Computer Intensive Learning EnvironmentS)
Version:1
Contact: Derek Rout
Date: 30 April 2010

         Also by using data gathered on rooms it can feed into a strategy to improve campus facilities
         when an opportunity is presented, leading towards a greener campus. The tool would be able
         to provide the data for energy saving strategies within the university. However this tool needs
         a lot more resources to show some real value. The tool highlighted some of the complicated
         data gathering required and needs considerable automation to show real time behaviour of
         spaces across campus. That real time data would then aid refinement in energy saving
         techniques.

        Making use of green forums, encouraging campus users to put forward energy saving ideas,
         and to debate the solutions suggested by their peers. If there are enough peers from different
         disciplines, the forum could be a good perception gathering portal, saving on trying to get the
         right people to attend a series of meeting to put their views across. Any campus user could
         access the forum and put their arguments forward, and would be encouraged to do so via
         champions and bulletins.

        The university will take into consideration developing induction programmes in order to
         convey a message of working greener, seeking ways of cutting back on energy waste and
         give practical applications of energy conservation. Some staff may have no idea of the new
         facilities available to them, to improve their carbon footprint.

        The university intends to embark on a Carbon Reduction Programme which will support and
         guide the university and provide a 5 year plan into reducing its carbon footprint.

        The work done by the ECCILES team has developed some changes in the university. The
         project has influenced the collaboration between the Facilities and IT departments and better
         understanding of each other‟s disciplines has emerged from having done such a project. Now
         when there are newer strategies proposed the departments are able to better co-ordinate the
         projects. The project team found the perception gathering exercise from TRIZ most
         influential, in understanding some of the barriers within the university infrastructure and
         culture which have influenced failed attempts to conserve energy in the past. The
         methodology seems to have brought together ideas and viewpoints and developed a strategy
         of action. The TRIZ approach gave every participant a say, it then relied on their input to
         progress with one strategy or the others'.

        The JISC funding gave the university an opportunity to devise a collaborative team and to
         look at the problem in greater detail. Without the interest JISC has had in the approach taken
         the project wouldn‟t have taken place in the way it did. The university would still be looking at
         energy savings at the Technical solutions level. JISC‟s interest in the application of TRIZ also
         helped the project remain on one track and benefit from the tools of the methodology. Given
         the number of influencing factors in the success or failure of applying any sort of a solution to
         the problem, TRIZ enabled the team to understand the background and users of the various
         areas. Why the end users are doing things the way they are, in order to design a solution to
         best fit all parties involved.

        ECCILES has tried to address the problem in all three levels Technical, Operational and
         organisational to highlight that saving energy is not simply a technical solution. Other
         universities can take any of the case studies from the project and implement them in their
         organisation or build upon any of the work carried out by ECCILES. The Ecciles team would
         be prepared to discuss any of the technical progresses of the case studies and support
         further applications of the ideas.

Recommendations (optional)
ECCILES would recommend using TRIZ as a problem solving tool where there are a lot of key players
who can influence the success of a solution. If for example when trying to implement energy saving
measures you come up against a lot of resistance because the people involved don‟t think your
solution is feasible then use a perception mapping exercise to get a better understanding as to why
you have that resistance. The benefit of using TRIZ it to reduce the failure of a solution system being
implemented. To get better Buy-in from all involved.



                                                  Page 36 of 37
Project Acronym: ECCILES (Energy Conservation in Computer Intensive Learning EnvironmentS)
Version:1
Contact: Derek Rout
Date: 30 April 2010

Awareness campaigns were received in different ways. The posters were not received as well as
planned but the Trip to Dinorwig was beneficial for the team in gathering end user perceptions and to
communicate what the ECCILES team was trying to do. When rolling out a solution to any large
areas of campus the team recommend that such an exercise be carried out half way through devising
a solution. This is to ensure you have a full understanding of the end users, (those you are trying to
influence) to discuss what you have planned and to gather response perceptions. As they are the
front line users (closest to the problem) they would be able to refine your solution better or even
highlight potential problems you may face in implementing your solution. It then gives you time to
adjust and amend the solution. The trip proved to an effective approach because the team could
capitalise on travel time to have effective debates.

When attempting to solve energy consumption problems the team recommend that a holistic and
collaborative approach is taken. Analyse what the aims and benefits from taking a specific approach.
For example the Ecciles team looked into sub-metering, and discussions went from data gathering to
empowering departments in being able to control their own power consumption. Making energy
another controllable resource in the same way as human, space or equipment resources are
managed.

Keep your target market and industry in mind when adopting a solution. ECCILES focussed its‟
efforts on Computer intensive environments where the industry is growing and developing
improvements very quickly. For example large Operationg system suppliers are writing in energy
saving measures into their operating systems. Which would make solutions such as power
management less effective. However power management enables the team to analyse the demand
on the equipment, understand peak usage and needs, enabling the university to make facilities
available upon high demand and have shut down strategies when demand is not very high.

References




<List any references to the work of others you have cited (e.g. articles, reports, studies, standards),
and any explanatory notes. Provide URLs for any materials available on the web.>


Appendixes (optional)
Appendix 1 – Triz Case Study
Appendix 2 – Metering Case Study
Appendix 3 – Lighting Case Study
Appendix 4 – Energy modelling Case Study
Appendix 5 – Awareness Case Study
Appendix 6 – Powerman Case Study
Appendix 7 – Virtualisation Case Study
Appendix 8 – Embedding Case Study




                                                  Page 37 of 37

				
DOCUMENT INFO