September 14th 2015

Building models, modelling buildings

Building models, modelling buildings

As a software developer in a building performance analytics firm, Michael Bennett uses his physics skills to help design more environmentally friendly and cost-effective buildings.

When architects and engineers design new buildings, they have a lot of different factors to consider. Lighting, shading, wind direction, heating, ventilation, airflow and many other elements all need to be taken into account. However, the rising costs of heating and cooling – coupled with concerns about climate change – mean that the way buildings use energy is also an important part of their design.

The company I work for, Integrated Environmental Solutions (IES), offers integrated software and consulting services that help architects, engineers and everyone else involved in the creation of a building to make better performing, sustainable and energy-efficient buildings. Our software analyses a number of different inputs (including climate data, building design, and the design of heating, ventilation and air-conditioning systems, among others) to calculate a building’s energy consumption and performance and suggest the best possible design strategies.

At first glance, it might seem strange that an astrophysicist like me would find a niche in this industry. However, in a nice coincidence, some of the heat processes (such as diffusion or convection) in stellar models that were relevant to my PhD are also relevant for buildings. In a few cases, the algorithms are directly applicable. For example, the finite difference methods used to calculate diffusion of heat across a wall can also be used to calculate matter transport in the interior of a star. I think it is amazing and wonderful how parallels can sometimes be drawn between celestial bodies and things closer to home.

Putting core knowledge to use

Following my PhD at the University of Keele I was unsure whether to pursue a career in academia or industry. I found doing research at the frontier of current knowledge exciting and interesting, and I was fascinated by the software models that are part of such research. However, I also wanted job security and the ability to settle somewhere and call it home. So, after some careful consideration, I started looking at industry.

This was in 2010, however, and with the UK economy in recession, there didn’t seem to be much demand for physics graduates. Most of my colleagues pursued academic careers, became physics teachers or found themselves in careers where physics was less relevant. Disheartened, I reluctantly concluded that in order to get a job, I would need to focus more on the value of my mathematical ability and transferable skills, rather than the value of the core physics knowledge I obtained during my degree.

The turning point came when, after several unsuccessful interviews with software and engineering companies, I signed up for recruitment agencies specializing in science and software related careers. Most of these agencies did not seem very motivated to find me work, but one of the exceptions, ECM, put me in touch with IES. When I found out that they were based in Glasgow, I was shocked, as I had mainly been looking for jobs near my friends and family in London. Nevertheless, I went for an interview and was happy to find that IES was looking for physics graduates to work on software related to heat transfer. It was clear from the interview that my core physics knowledge would be valued and so, after some thought, I accepted the job and moved to Scotland.

A model assignment

Although I am a software engineer, almost all of my daily tasks require physics knowledge and skills. Before I can start designing software, for example, I need to understand the model that is being implemented and identify its limitations and capabilities. This requires me to retrieve, understand and critique scientific literature – a task that often entails a considerable amount of mathematics, especially when the model involves fluid transport (usually air or water, but sometimes refrigerant or other substances), heat exchangers, convection or solar radiation.

Once I am confident that I understand the technologies and processes I am modelling, my next task is to develop proof-of-concept models, so that I can identify possible complications or unusual situations that might arise. For example, certain renewable technologies involve convective heat transfer from a surface into an air stream at a flow rate relative to some design flow rate for the system. What if a user specifies a very low design flow rate or doesn’t specify one at all? We will need to consider natural convection, transition to turbulence and sensible default values for flow rates in the absence of important input data. Preliminary models of such things can identify issues such as these.

The next step is to write (and then test) the software code. I also perform validation studies in which I compare the output of our thermal models with real building data. Sometimes I perform uncertainty studies, too, to check how robust the models are.

Coding and communication

Because my role is “client-facing”, I regularly keep our clients updated on progress, and I also give presentations regarding the outcomes of our studies. Communication skills are essential as we often work with people who have had little or no exposure to analytical and numerical models of physical phenomena (especially the physics of buildings), and who therefore find such models incredibly complicated. Being able to break complex scenarios down into something that is simple to understand, and then build on that in order to describe the situation more fully, is very important.

As for the coding part of my job, I learned most of the basics required to write software code during the last year of my physics undergraduate degree. However, I also developed these skills considerably during my postgraduate studies and in my spare time, so if you are interested in a career in software engineering, I recommend spending some of your free time programming. There are many amazing open-source resources on the Internet and tutorial websites that can help you learn, and if you choose a project that sounds fun, you will be more motivated to continue working on it. For example, you could try making a simple computer game or a “physics sandbox” tool in a high-level language (like Python or Lua). Then, if you want more of a challenge, you can go on to experiment with lower-level languages such as C++ or Java. There is a huge demand for programming skills at the moment and they are incredibly useful for many things (including science).

While communication skills are important, you don’t have to be an amazing performance artist or conversationalist to be able to communicate your ideas effectively. If communication isn’t your strong point, practise! Give short presentations to others, form discussion groups, talk about technical topics with people you know and gauge their reactions. Don’t be afraid to ask for feedback so you can identify areas where you can improve. There are plenty of opportunities to do this during a physics degree and it can even be fun if you’re discussing something you are passionate about. It will also help you with job interviews.

My advice is to be active, open and motivated. I’ve found that the best aspect of my job is the knowledge that my experiences, science knowledge and skills are useful to people. After spending many years studying physics and mathematics, it is flattering when people need my help and rewarding when I can see how their jobs are made easier following my input. It is also great to be able to continue studying physics and learning new skills as I do my job; nothing feels static or predictable and the projects are always varied and intellectually stimulating. My experience shows that there are great physics jobs out there if you look for them.

Michael Bennett is a software engineer at Integrated Environmental Solutions in Glasgow, UK. For information about vacancies, visit www.iesve.com/jobs or follow @IESCareers on Twitter

This article was originally published in the September 2015 issue of Physics World Magazine, http://physicsworld.com/.