Information Technology Reference
In-Depth Information
-
All processors seek suppliers. Depending on the scale of operation they preferably
use the intermediaries - true for all processors equal or bigger than anaerobic di-
gestors.
-
In network terms the emerging process might be seen as a diffusion of information
(price, quality and quantity). Agents operating in the waste context would be repre-
sented as a serial diffusion process (perhaps with some kind of learning). By contrast
the diffusion of information of consultants might be represented as a parallel process
in which agents basically have access to the same information (prices in a global mar-
ket). This leaves us with the oil brokers. Unlike facilitators and brokers they could
behave strategically and hold the flow (subject to long-term supply contracts) since
they own the resource.
7
Discussion
The ERIE project progressed in two phases. Both relied on qualitative methods to de-
fine rules for the agents of an agent-based model. Despite this similarity we consider the
two phases as distinct. The first phase consisting of semi-structured interviews effec-
tively enabled us to understand relevant aspects of the developing bio-based economy
in the Humber area. Whereas we had a clearly defined theoretical framework and col-
lected sufficient data from the standpoint of qualitative social research we did not obtain
enough data to create a defensible and relevant agent-based model. Instead we had to
acknowledge that networks and institutions are empirically more complex than cogni-
tion resulting in an under-defined toy model. This led to a second phase in which we
redesigned our conceptual model according to empirical problem observation and eco-
nomic literature. We entered a third phase where we engage in a participatory modelling
process, which is on-going. One of the reasons for choosing a participatory method over
conventional qualitative methods such as interviews is the higher information density
achievable with participatory exercises.
Once implemented in an agent-based model the empirically established rules should
already result in complex dynamics such as market price fluctuations. The aim is to
detect price volatility emerging from these interactions and mitigating effects resulting
from e.g. long-term contracts or government policies. Price volatility is a widely recog-
nised problem for companies and may become more pertinent in future as some studies
suggest a strong relationship between bio energy and oil prices [10]. Furthermore, price
volatility is a particular problem of secondary material (recyclate) markets more gen-
erally because these markets often lack market clearance of waste outputs according
to supply and demand [1,2,4]. For this reason the UK government has endorsed the
creation of recyclate markets from early on [15] by creating for instance one of the fa-
cilitators (NISP) we saw in the area network. This might be of less concern in respect to
organic waste as it can either be recycled to land (source-segregated organic waste) or
brown field sites (co-mingled organic waste). There is thus a host of questions around
the evolution of the bio-based economic system we can now explore with the revised
agent-based model.
We want to pursue three research questions in particular: firstly, we want to under-
stand path dependency in the evolution of the bio-based economy in the Humber area.
