Cookaβurra '11

Case from Deltares

Two Phase Flow in a Declining Pipe

Students

• Hans Kuipers
• Geert Reitsma

Supervisor

A.E.P. Veldman

Contact

I. Pothof

Since January 1st 2008, the Netherlands benefits from a new and independent institute for applied research and specialist advice. Together with Rijkswaterstaat/DWW, RIKZ and RIZA, WL | Delft Hydraulics, GeoDelft, and a part of TNO Built Environment and Geosciences form the Deltares Institute. The institute employs more than 800 people. Deltares has a unique combination of knowledge and experience in the field of water, soil and the subsurface. It is frontrunner in the development, distribution and application of knowledge for meeting the challenges in the physical planning, design and management of vulnerable deltas, coastal areas and river basins. Deltares works for and cooperates with the Dutch government, provinces and water boards, international governments, knowledge institutes and market parties. The institute is located in two cities: Delft and Utrecht.

Background of the case-study

Gas pockets in pressurised waste water mains cause significant capacity re- ductions, resulting in unnecessary CSOs (Combined Sewer Overflows) and excessive power input. The current state of knowledge on the rate at which a gas pocket is transported through a downward slope, is limited. The CAPWAT joint industry project (Capacity reduction in wastewater pressure mains) is co-funded by most Dutch water boards, several consultants and research foundations. The objectives of the CAPWAT project include:

• Development of a detection method for gas pockets;
• Development of measures to minimise air inflow;
• Determination of the required velocity to remove capacity reducing gas pockets;
• Rate of gas pocket breakdown and transport;
• Guidelines for design, operation and maintenance of wastewater mains.

High speed camera observations have been made in order to get information about behaviour of gas pockets in pressurised water.

The case study

A certain amount of gas and a certain amount of water are flown through a declining pipe. Thirteen series, each series with other flow rates, of pictures are taken from above. Each series contains 100 pictures. The pictures show a pattern of bubbles. From this pattern the drag coefficient has to be estimated.

Method

A formula for the drag coefficient is derived by using archimedes law and the formula for the drag forces.

• A_i the interface area of the bubble; this quantity is determined directly from the TIFFs;
• α the angle of the pipe;
• ρ the mass density of the fluid; In this case 1000 g/l;
• v the speed of the bubble relative to the fluid; The speed of the bubble can be determined directly from the TIFFs, the speed of the fluid can be calculated if the outflow and the cross-section of the pipe are known;
• C_d the drag coefficient of the bubble;
• A_b the frontal (or bubble) area of the bubble; This quantity is obtained by measuring the maximal width of the bubble, multiplied by a constant h;
• h is the average height of the bubbles, which can be calculated if the total outflow of air and the number of bubbles per second is known. We assume here that the difference in bubble heights is negligibly small.
• dH/ds is the pressure gradient of the liquid in the declining pipe. The pressure gradient is determined by using the Colebrook-White equation and the Swamee-Jain equation.

A picture viewer and a picture manipulation program are used to determine bubble velocity, interfacial area of each bubble and the maximum width of each bubble.

Results

The relations between drag coefficient and bubble volume, between drag coefficient and fluid velocity, between drag coefficient and distance between bubbles and between drag coefficient and relative velocity are obtained. The relation between drag coefficient and bubble volume strongly depends on the ratio between gas flow rate and liquid flow rate. The drag coefficient seems to be constant for different fluid velocities in the range of 100 – 150 cm/s. Also the distance between the bubbles does not influence the drag coefficient.

Case from Force Vision

Students

• Mark Abspoel
• Mark IJbema

Supervisors

• F. Brokken
• J. Top

Contact

J.M. de Vries

Force Vision is the software engineering department of the Defence Material Organization (DMO). The company designs, develops and implements combat management systems for the Navy’s frigates and submarines. A combat management system forms the command and control core of a ship, integrating the various sensor-, weapon-, and control systems used on board. In business since 1967, Force Vision is one of the oldest software engineering companies in The Netherlands.

The security of information is crucial for a combat ship. Classified information should only be available to authorized personnel, and the integrity of the data should remain intact. Also, information should be available directly whenever it is needed. This summarizes the three basic principles of information security: confidentiality, integrity and availability. It is a complex task to check whether a large information system satisfies these basic principles.

This process, called a computer security audit, involves mapping out all entities and information flows within the system and check where breaches of the security might occur. To assist in such an audit, risk analysis methods and tools have been developed. These provide a structural approach for mapping out an entire system, and supply lists of possible security breaches for each type of entity within the system. The tools also provide possible countermeasures to ensure the safety of the system. The Dutch Ministry of Defense prescribes the usage of the risk analysis and -management tool CRAMM for all its departments. This tool is very general however, developed for use with a wide spectrum of systems, and sometimes offers suggestions inapplicable on board a naval vessel. Force Vision therefore asked us to find out which conditions a risk analysis tool should satisfy to be usable for auditing the information security of their systems.

Because we had no access to CRAMM itself we informed ourselves about the system by reading the manuals as well as independent papers and reviews of CRAMM and comparable systems. Furthermore, we reviewed some articles on the usage of CRAMM within the Dutch Ministry of Defense. With the knowledge obtained from these texts we researched the possible demands on such a risk analysis tool.

We eventually produced a list of conditions a risk analysis tool should satisfy. Firstly, the tool should comply to well-tried security principles and conform to the rules and regulations of the organization. Secondly, the situation of Force Vision implies more specific demands on the tool. Specifically it should be able to handle real-time availability and non-monetary risks. More generally it is required that the tool suggests relevant controls and allows for modification of controls and risks. Since we did not know all the specifics of the systems of Force Vision, we additionally suggested a test which should determine whether a certain tool would be applicable to their situation. This test consists of constructing several small test cases, and consequently modeling these cases with the tool, verifying the results and usability of the tool.

Case from GasTerra

Students

• Jelle Blijleven
• Roel Tempelaar

Supervisor

T. Schoot Uiterkamp

Contact

J. Hoogakker

Renewable energy resources are expected to have an even greater impact on the energy market in the next decades than is already the case. Natural reserves of fossil fuels are limited and as the global energy consumption keeps increasing over time, renewables will become the new standard. This transition will be accompanied by a problem of capacity. Since renewables like solar and wind are dependent on the weather, a constant capacity cannot be assured by these renewables alone. These unwanted but inevitable variations of capacity have a direct in?uence on the organization of the energy market. They must be taken care of, lest power delivery is endangered or even blackouts show up. This is where we enter the stage.

GasTerra a is company trading natural gas on a large scale. For obvious reasons, GasTerra has great interest in the development of the availability of natural gas in the future. Our mission was to investigate the possibilities for natural gas to solve the beforementioned capacity problem, with the goal to present to GasTerra a proposal for further research.

As a part of our investigation, we met with three experts and talked about the ins and outs of the energy market and the future outsights. Catrinus Jepma, president of the Energy Delta Convention, was the ?rst candidate, followed by Fons van Dam at the Nederlandse Gasunie, and Aad Correljé, of Delft University of Technology. We obtained a lot of information during these interview sessions. The rest of the study was performed using de?nitions and information, gained from the scienti?c and so-called ‘grey’ literature, and the press.

The research proposal was outlined in an extensive report. A plan was presented to investigate the capacity problem in north-western Europe during the period up until 2030. A short overview of the most important energy market factors was given, accompanied by a research model. The report was handed over to GasTerra, where hopefully the research proposal will be executed in the near future.

We really enjoyed investigating this subject. It was a great opportunity to dive into the complicated world of energy, a branch of research that lies on the border between economic, political and scienti?c issues, which makes it so interesting. The capacity problems will become reality and we eagerly await the political and institutional decisions to counter them.

Case from NAM

Students

• Morten Bakker
• Robert Broos

Supervisor

L.P.B.M. Janssen

Contacts

• R. Rijpkema
• R. Smit

The demand for natural gas strongly depends on the weather. During warm summer days it is almost zero and during cold winter days it is very high. Most of this demand is met by producing natural gas from the large Slochteren gas field in Groningen by production sites called clusters. However, there is a maximum volume that can be produced, therefore the NAM uses so called underground gas storage facilities (UGS), which are empty gas fields that can be refilled to meet the excess in demand during cold winter days. These UGS reservoirs act as a buffer but need to be refilled after use. This is done in the summer when the market demand for natural gas is low.

But the UGS is not only used as a buffer. It can also be used to minimize the power usage during gas production. This works as follows. During low gas demand the cluster can operate in so called ‘free flow’, meaning that pressure in the field is high enough to produce the volume that is needed. For higher demand however, compressors are used to get more gas out of the reservoir. The required energy to produce a certain flow of gas increases exponentially (see figure 1), making it very costly to produce gas at high volumes. This is not the case for the UGS, which has a large free flow volume. However, refilling the UGS does require energy.

For low volumes of demand the cost to refill the UGS is usually higher than the cost to use the cluster at an increased flow. But if the demand (in winter) becomes higher than a certain volume it can become more efficient to produce from the UGS in free flow and re-inject this gas in the summer. This leads to a so called ‘x MW switching rule’, meaning it is more energy efficient to switch over to production from the UGS when the power usage of the clusters increases above x MW.

Minimizing the power usage saves money and is good for the environment. But we can take this optimization one step further. Our case study was to extend the switching rule to also include electricity cost. The total cost of the required energy to produce gas does not only depend on the power required but also on when it is required: the energy cost is higher during the daytime than at night or during the weekend. But more factors play a role, such as the total UGS volume that is available and the contracts with the electricity companies. We made a model that incorporates these factors both for the coming years as on the long term.

Our main findings were that it is in principle possible to save more on electricity cost and energy by making smart use of the UGS. However, this will become more difficult in the future when the Slochteren gas field gets depleted and the costs to re-inject gas increase. Our simulation helped gaining a greater understanding of the possibilities to use the UGS for energy saving and is a good starting point for the NAM in developing successful strategies.

Case from Noordhoff Uitgevers, Moderne Wiskunde

Students

• Erik Duisterwinkel
• Auke Sytema

Supervisor

G. Vegter

Contact

S. Jissink

About Noordhoff Uitgevers

Noordhoff Uitgevers (Noordhoff Publishers) has its roots in Groningen. In 1836, Jan Berends Wolters started a book- and papershop at the Guldenstraat. This shop expanded to become a renowned publisher of books like the Bos’ schoolatlas (from 1877), the novels about Ot and Sien (1902) and the famous Aap-Noot-Mies (about 1910). In 1858 Popko Noordhoff started a publishing office at the Herestraat. He specialised in scientific and school books.

In 1968 Wolters and Noordhoff merged. Until 2007 Wolters-Noordhoff was part of the division of Education at Wolters Kluwer, then they were sold to Bridge-point Capital. This caused the change of name from “Wolters Noordhoff” to “Noordhoff Uitgevers”.

Since its beginning as Wolters in the Guldenstraat, Noordhoff Uitgevers has developed to become the largest educative publisher of the Netherlands, with over 400 employees today, established in Groningen and Houten.

Moderne Wiskunde

During math classes in high school, a much-used method is the book “Moderne Wiskunde” (“Modern Mathematics”). This book is accompanied by an electronic study environment on the website of Noordhoff. This site contains a number of i-clips where theory is explained and exercises can be made. For this casestudy, Noordhoff had the request to create a new i-clip, meant to practice calculation skills. The concept was to practice mental calculation with the operations like add, subtract, multiply, divide, take the root and raise to some power. We made a game concept in which the problems and answers exist as balls that roll over the game field and need to be combined with each other in the right way.

The making of the game

The game (see figure) was made with Adobe Flash CS4, with the code written in ActionScript 3.0. First, we made a sum generator that creates an arbitrary problem with some operation and criteria concerning the numbers. For example, with multiplication the first number contains at most two digits and the second number at most one.

A game field with sumballs (balls with problems) was made, and at fixed moments new sumballs are generated at the top left corner of the screen. When an answer is typed and the enter button is pressed, the answerball appears with the – hopefully correct – answer. The sum- and answerballs can be dragged by mouse, so that they can be combined. And when they get impulse by using the mouse they will keep rolling until they hit each other. When the answer is correct the player is rewarded points and otherwise he or she gets a penalty.

A collision algorithm was programmed that enables the balls to collide realistically. Also, a saw was added that can remove answerballs, and the sumballs explode after some time creating a shockwave. A scoreboard and a level setup with five levels and three difficulty levels was made.

The game has been sent to the authors of the Moderne Wiskunde manual and it was received well. There were some ideas for further improvement and for features that could be added. The current version will be put on the website of Noordhoff, and maybe a new version will be made in the future.

Case from SRON

Students

• Monique Ankoné
• Jasper van Dijk

Supervisor

M. Aiello

Contact

P. Roelfsema

SRON, Dutch Institute for Space Research, is a Dutch agency that was founded in 1983. The institute has two facilities, one is located in Utrecht and the other one in Groningen.

SRON focuses her attention on the development of satellite instruments. The institute develops and uses innovative technology for groundbreaking research in space focusing on astrophysical research, earth sciences and planetary research. In addition to this, SRON has a line of research into new and more sensitive sensors for X-rays and infrared radiation. The selection of these research disciplines is based on choices made on the basis of instrumental expertise and on the ambition to act as Principal Investigator (PI) in a few preselected science areas.

As a part of the Dutch Organisation for Scientific Research, SRON is the national center of expertise for the development and exploitation of satellite instruments in astrophysics and earth system science. It acts as the Dutch national agency for space research and as the national point of contact for ESA programs.

SRON provides ‘the ensemble of knowledge and skills, both technically and scientifically, required to perform a principal role in the scientific utilization of space.’

The casestudy

Observations in the infrared, sub-millimeter and millimeter windows are of great importance in astronomy. It is therefore that the European Space Agency (ESA) built and launched (on May 14th 2009) the Herschel satellite, to explore our universe in those wavelengths.

SRON is the Principal Investigator Institute for one of the three instruments of Herschel, the Heterodyne Instrument for the Far Infrared (HIFI). HIFI will provide us a continuous coverage over the range of 480 to 1250 GHz in five bands and over the range of 1410 to 1910 GHz in two additional bands. However the information of this instrument is spread out over three different webpages.

• sron.rug.nl/hifi_icc
• sron.nl/divisions/lea/hifi
• sron.nl -> divisions -> low energy astrophysics -> hifi instr. / hifi science

It was our assignment to create one website for the HIFI instrument. One website to contain all the information of the three websites, displayed in a structured way. To structure the website we talked to many people from SRON and others (i.e. scientists and students), who could give us more insight in how the website should be structured. This enabled us to create a tree that resembled the structure of the new website. The final result will be shown at sron.nl/divisons/lea/hifi.

Case from The New Voting Foundation

Students

• Sander Land
• Maarten Kruijver

Contact

S. van Grieken

The New Voting Foundation (Stichting Het Nieuwe Stemmen) aims to stimulate the participation of ordinary citizens in politics and public office. The foundation tries to achieve this goal through harnessing the power of the internet to bring politics closer to citizens, whom it’s all about. The foundation consists of a bunch of nerds who get extatic about developing new groundbreaking applications that can gather and present political information in ways never before possible.

In practice, this goal of bring together politicians and citizens is achieved by creating websites containing for example voting guides or information about politicians. These websites used to be ad hoc solutions to gaps in the communication of politicians and citizens. A new project of the foundation is however posed as a more fundamental answer to the problem of citizens not feeling connected to decision makers. This new project is called You and the Government (jijendeoverheid) and consists of one single website on which the government communicates with her civilians.

The case study touches upon one aspect of this You and the Government website project, specifically the aspect of frameworks. An important part of the project is the so called hns.dev database, which lies at the heart of the new central governmental website. This custom built database has several technical features that make it possible to serve millions of pages a day, while maintaining a high level of security with different access roles. For the system to function there has to be put a framework on top of this database. Our case study consisted of finding out which framework is suitable for this daunting task.

As a start, we had to find out what the foundation expected from the framework. It turned out that they expected an extremely flexible and reliable open source framework capable of interacting with the xml queries and responses of the hns.dev database. Additionally, it was a wish of the foundation to be able to embed framework widgets into external websites. Based on the wish list we came up with a list of requirements for the framework.

The next step in solving the case problem was to identify several target frameworks which would potentially be usable. We came up with a shortlist of candidates. After this, we decided to compare the candidates on several aspects, making it possible to make a final decision about the technology to implement. Yet, the project is still ongoing, but we expect to see updates from the foundation in the near future!

Case from the University Library

Student

• Thomas ten Cate

Supervisor

H. Ellerman

The University Library at the University of Groningen is the university’s main library. Not only does it offer a physical location (the kind with books) across the street from the Academy Building, it also maintains the various digital catalogues and search engines that are in use throughout the university.

My casestudy focused on one of these search engines, PurpleSearch, which can be used to search for books, articles and other publications. PurpleSearch is a so-called meta search engine, which means that it does not have a da-tabase of its own, but instead fowards the search request to various other University Library catalogues. Because it has access to nearly a hundred different catalogues, PurpleSearch contains some “intelligence” to select a subset of these catalogues that are most likely to return useful results for this particular search query. The actual search queries from PurpleSearch are executed by the software MetaLib by Ex Libris. MetaLib offers the possibility to write arbitrary programs to perform searches, which is what my casestudy was involved with.

Although MetaLib ships with a respectable number of such external scripts, all of these are Perl code of the hairy variety. The developers of PurpleSearch desired a more readable and maintainable solu-tion, written in Python. My task, then, was to write a Python program to perform searches on behalf of MetaLib on servers that use the SRU protocol. SRU stands for “Search/Retrieval via URL”, and is a standardized interface that can be used to perform a search query using a single HTTP request, returning the search results in an XML-based format.

The Python code that I wrote for this task essentially consists of three parts. On the one side, there is the interface with MetaLib. The program needs to parse the request that comes in from MetaLib and after the search, format a response to send back to MetaLib. On the other side, there is the SRU interface. The program needs to send an HTTP request and process the resulting XML response. In between these two is a conversion step, that takes the incoming search request from MetaLib and converts it to SRU format. An important part of this converter is the query processor, which converts the actual search query from MetaLib’s internal format into CQL, the Contextual Query Language, which is a part of the SRU specification. Another important part is the conversion from the server’s XML response, which contains MARC (Machine Readable Catalogueing) XML records, into a plain-text MARC format that MetaLib understands. Because all these conversion steps can introduce errors, I wrote unit tests for all modules to ensure a certain degree of correctness.

The resulting Python code has been tested against two different SRU servers, namely WorldCat and the University’s OPC catalogue, and works quite well. MetaLib’s internal query format is hardly docu-mented, and the user’s search terms are inserted without escaping into a query, which made correct parsing impossible and robust parsing a challenge. However, the new parser is much more robust than the original Perl version. As it turned out, not all SRU servers conform precisely to the standard. To accommodate this, the package provides hooks on which custom code can be attached that works around the quirks of one particular server. The package thus makes it easy to add any SRU server to MetaLib, and forms a basis for implementing protocols other than SRU with little effort.

Case from UOCG (I)

Creating test cases for UOCG

Student

Gjalt Bearda

Supervisor

L. van der Duim

UOCG: Creating test cases The short description for this case study was twofold: first designing and making test case files for the interactions between the website Prog-ress.net and the website Studielink, and secondly improving the auto-matic test run program.

When constructing new software, you go through several development steps adding one function at a time. It is wise to make test cases and to test new functionalities. Unfortunately, programmers often forget this. They like to add new functionalities and neglect the need for good test cases in order to test what they have developed. You could say that they are too confident about the quality of their own products, however every new release of software always has some bugs.

The University Centre for Learning & Teaching (UCLT, in Dutch known as ‘UOCG’) is a Centre of the University of Groningen that has three main parts.

• Training teachers;
• Enhancing the quality of education, for example by quality assurance;
• Deployment and development of ICT products, with the emphasis on innovation in education

One of the products developed is Progress.net, which handles student administration from enrollments to grading. The Dutch government has developed the website Studielink where students must enroll for their programs. Since Progress.net does the student administration for the University of Groningen these two systems have to be linked.

I have designed test cases for another program developed at UCLT, the automatic test tool. This test tool has to be filled with test cases which have to run before a new version of Progress.net is released. The test cases exist of a sequence of tasks to be done automatically, like navigating to a webpage, clicking on a link on a page, filling in a form on a page or asserting that some text is on a page.

With combinations of these tasks the functionality of the new version can be tested. For example, changing your name at Studielink is done by completing a form. Then Studielink has to communicate to Progress.net that the name has changed (which is part of the functionality). Next the test program has to check whether the name indeed has changed in Progress.net by asserting the new name in the personality page in Progress.net.

The test cases tested all communications found between Studielink and Progress.net and back. In total four faults were found in the then latest version. Also some inconsistencies between Progress.net and Studielink were found, like the number of addresses.

Additional to the testing, some improvements for the automatic test tool were implemented. These improvements made it easier to build the test cases.

I would like to thank Ceesjan Luiten for implementing the improvements I asked for in the automatic test tool.

Case from UOCG (II)

Information Flows

Students

• Harm Jan Prins
• Femke van Seijen

Supervisor

L. van der Duim

The UOCG (in English UCLT: University Centre for Learning and Teach-ing) is an institute associated to the university of Groningen which pro-vides training of secondary school teachers and administrates several systems which are used by the university for educational purposes. For this casestudy we investigated the flow of information between these different sys-tems. Examples of these systems are systems for administration of students, teachers, courses and marks, or an online interactive educational platform.

The first thing we did was narrowing our field of research, because there were more systems then we could analyse in the given time. To find out what kind of information is transferred, we talked with experts of the different systems. They gave us a lot of useful information which we next put in a clear and trans-parent form. We did that by making models of the databases and diagrams of the information flow. This was done on a semantic level, we focused on the meaning of the information, not on the way it was represented. By construct-ing the models questions arose with which we again consulted the experts to improve our model; this was done repeatedly.

In the end we had an eighty pages long report with a lot of diagrams and models which gave an overview of the information flows between the five most impor-tant systems of the UOCG. Our main conclusion is that every expert knows how his own system works and what kind of information is exchanged with other systems, but that nobody had the total overview. With our report we delivered this necessary overview. Minor conclusions were about particular systems. These systems were developed in a long period, but sometimes decisions were made based on short term efficiency instead of long term clearness. These decisions had as consequence that certain information in these systems took strange paths before ending up where it was needed.

We also made some recommendations how to improve the information flows. The main recommendation was to choose one system to be the central system with attached to it the central database. All new information then has to be put in this system before it can be used by other systems. In this way a lot of prob-lems can be omitted, because it is clear where to get your information. This prevents that different systems say different things about the same entity.

Case from Science Shop Mathematics and Natural Sciences

Student

Rob Bremer

Supervisor

R.E.M. Neubert

The Science Shop Mathematics and Natural Sciences does research for public organisations, like foundations and associations on the field of chemical, environmental and labour conditions. The aim is to avoid financial barriers for the initiator. It is even possible for organisations to get researches for free, if they are running low in funds for their research. All researches are carried out by students who can use it for a course, bachelor- or master research.

The research of this case study was assigned by the province Friesland. The aim of the research was to measure the average sound level of the surrounding noise in silence areas in the national parks of Friesland. Research is carried out in ‘NP het Drents Friese wold’, ‘NP de Alde Feanen’, ‘Ameland’, ‘NP Schiermonnikoog’ and ‘Vlieland’. Each location is monitored during two days with a decibel meter. During the measurements, with the help of a stopwatch, all sounds are identified and registered. For example, if from the time 1:15 min. until 1:45 min. a car is heard, this will be registered. The measured data will be analysed with the help of these notes. The total time of each sound source can be determined. From this analysis the contribution of each source to the total sound level can be expressed in percentages.

The aim of this research is to monitor the silence areas, and investigate what are the main not natural sources that give rise to the background sound level. Also statistical values of the background sound are determined, like the L95 value which is the sound level with only the 5% of the lowest sounds. This is a measurement for the characteristic properties of a location. This value is only influenced by the sound of the wind and other natural sounds. Also the Leq value is calculated, which is the average during a complete measurement. Because all values are measured in decibels the loud sounds have a higher weight factor and will influence the Leq values more than other sounds. At each location in a park the difference between the L15 and Leq value gives a measure for the amount of disturbance, where the L15 value is all the surrounding sound minus the top 15% of the loudest noises.

All these results are compared with the results of a report from 2005, with the exact same research, also performed by a student from the University of Groningen. For a fair comparison, the results will be corrected for the differences in the sound of the wind and other ‘natural’ sounds, by looking at the L95 values. The first impressions: the background levels are almost the same, in some parks there are less disturbances. The final results will be presented in a report and in a presentation in the presence of all the local governments and some people of the County Council.