Help System

How to Use the Web Model

The web model is designed to allow visitors to the site to interact with the ECCO model. The help section here describes:

• What a simulation model is

• Why use the simulator as a policy-testing tool

• How to drive the ECCO model simulator

A simulation model is a tool for understanding how something works. Different parts of the 'something' are represented in the model, along with the relationships between them. By formalising all the parts and their relationships, hopefully one can spot connections that wouldn't be obvious in an unstructured or 'back of the envelope' analysis.

A system dynamics model (the family to which ECCO belongs) uses numerical equations to represent all parts of the 'something', and presents results as time series data, commonly displayed as graphs.

System dynamics models are generally open-ended (ECCO certainly is). We do not specify what the end-point of the simulation is, how the 'story' told by the model will end. Rather we set up the initial conditions, and then let the story unfold (sometimes with a few pre-programmed interventions). When using a simulation model, there are typically three broad phases to follow.

• set up the starting conditions, or scenario

• run the model

• interpret the results

In extended use of the model, these three steps will be repeated many times, giving a set of results from many scenarios, which can be compared against one another.

Why use a simulation model?

There are pros and cons to the structured mathematical approach. On the plus side, non-intuitive behaviour can be uncovered, and new insights into the situation being modelled can be developed. On the minus side, it is impossible to characterise any situation completely in the formal sense required by a simulation model. A model of a complex situation can quickly become very large, yet still contain approximations that seriously limit its use. When interpreting the result of a model, it is necessary to bear these approximations and limitations in mind. The danger is that the model may be too detailed to hold all the limitations in mind easily when interpreting.

ECCO attempts to provide the minimum of detail required to capture the behaviour of a national economy (despite which a mature model will contain some 2000 equations!). We are more interested in the patterns of the behaviour generated than the actual numbers as hard and fast predictions, although we strive to make the behaviour of the base case scenarios match up to reality as accurately as possible.

Because a simulation model holds information on many parts of its subject at once, it can be a good tool for holistic thinking, that is, thinking about the interconnectedness of the subject as a whole. With ECCO, we hope to provide fresh insights into the interconnectedness of our society, for example:

• what effect would a shift towards renewable (or nuclear, or fossil) electricity generation have on consumer spending and affluence?

• how would large-scale investment in biomass energy impact long-term economic growth?

• how fast would we need to invest in renewable energy technologies in order to maintain domestic energy supplies if oil supplies were to be exhausted in 15 years? in 50 years? in 100 years?

• what impact would increased foreign investments have on domestic carbon dioxide emissions?

The above list is simply a few examples. Any combination of policies can be combined for a given run. ECCO models have policy options for a wide range of issues, including energy resources, electricity generation technologies, agriculture, investment, international finance, employment, population, education, transport, housing, pollution and waste management, resource efficiency, to name a few.

These questions are cross-sectoral. Traditional qualitative policy analysis has often restricted itself to a single issue, because of the enormity of the job of mentally holding all the cross-sectoral impacts and ramifications. It is here that we believe ECCO's great strength lies, in complementing and enriching traditional policy analysis with a rounded holistic overview.

How to drive the ECCO Simulator

The ECCO modelling software that can be used in this website or downloaded for offline use follows the three-phase pattern described above, common to all simulation modelling. To recap:

• set up the starting conditions, or scenario

• run the model

• interpret the results

The main simulator runs in its own window, and can be launched from the startup button on the title bar of this (and every) page of the site.

Because the web model is quite large to download (around 400kbytes), and because it relies on technologies that not every web browser will support (although most do), it is not automatically started up when this site is first visited.

It is possible to tell if the simulator has already begun downloading to your computer because the title frame for the site will display a startup button with a message that changes as simulation proceeds (see figures 1 & 2 below). If the titlebar is empty, the download has not yet begun, and may be started by clicking here. As the simulator downloads the components it needs, the 'traffic light' icons will fill in (see figure 2).

It is save to browse around the text portions of the site while the model loads up without interrupting the simulation.

The model does not need to be run from a live internet connection. All the components necessary for running the model offline can be downloaded from here.

Figure 1: Title bar without model loading in the background - click on the icon shown to begin loading

Figure 2: Title bar with model loading in the background - the traffic light icons and text message show the progress of the download.

Once the green light is visible (the third state shown in figure 2), the simulator application can be launched by clicking anywhere on the startup button. It pops up in a separate window as shown in figure 3.

Figure 3: simulator appears in a separate window

The application window has five main 'panels', each accessed by clicking on the appropriate tab at the bottom of the window. These are:

• Changes, which allows the user to set the values of the starting conditions, either from a pre-selected set of scenarios or one by one

• Save, which allows the user to select values to be saved in the result set of a given scenario, again from preselected groups or one by one.

• Go, which has controls for launching the simulation run, and managing multiple runs.

• Data and Graph, which allow the user to view and compare stored results, as text data and graphs respectively, and to export data into spreadsheet programs for more detailed analysis.

More details on the operation of each of these panels can be found by clicking on the links above.