The Changing World of Europe's Asstes

The Changing World of Europe's Asstes

Ben Rodgers
GIS Analyst, Petroleum Services
Deloitte & Touche, 180 Strand, London, WC2R 1BL, United Kingdom
Telephone: +44 (0) 207 438 2941
Fax: +44 (0) 207 438 3881
Email: brodgers@deloitte.co.uk

Abstract
As European gas markets become increasingly more connected both physically and contractually, gas assets will need to rise to the challenge. Some assets will find that their location in the new gas market is more (or less) advantageous than it was in the past. To assist companies in removing this uncertainty and to help formulate plans for the future the ‘gas toolbox’ was developed. The toolbox brings together asset economics, spatial data and GIS technology to create credible scenarios of European gas supply and demand. Asset based forecasting models are first developed in Microsoft Excel, allowing users to speculate about the supply and demand characteristics of gas assets and to provide risk frameworks in decision support. Once loaded into our ArcGIS system, users can visualise and add or remove assets from their scenarios. Developed using ArcObjects and Visual Basic, the toolbox has been customised to be a spatial and temporal model. Up to 20 years worth of data can be visualised and edited for each scenario. Individual costings of the gas transit routes from well head to burner tip are exposed by performing network analyses, taking into account planned, decommissioned or new assets for the selected year.

Background
As European gas markets become increasingly connected both physically and contractually, gas assets, gas players and the institutions that affect gas markets will need to rise to the challenge. The location of some assets in the new gas market will be more or less advantageous than in the past. Gas asset economics will be increasingly exposed to market forces and regulatory scrutiny. Asset operations, flexibility and option value may radically change.

During the next two decades the European gas market will play a pivotal role in Europe’s energy market and economy. Gas suppliers into Europe will play an expanded role within the wider global gas market place. The development of the European gas market could set the overall conditions for the global gas market.

Europe is uniquely positioned to capture maximum value from its location, gas infrastructure and diverse gas supplies. Gas will play a more important role in the European energy market in the future than it does today. European energy demand is diverse with oil, gas, hydro and coal all utilised. Europe, however, is import-dependent for gas supplies and takes delivery of large tranches of gas from Algeria, Norway and Russia. Whilst diverse sources of supply have been secured, Europe will always be susceptible to the overall market conditions of the global gas market place.

The conditions that shape the development of the physical gas market in Europe will have a major effect on the conditions of the whole European energy market.

With this in mind, Petroleum Services set out to develop a framework and then capture and model these conditions within a GIS. It was felt that much of the uncertainty in the European energy market is caused by a poor understanding of the energy infrastructure, such as the asset economics, demand and supply characteristics and gas network connectivity. Typically, GIS have been developed to model either the upstream or the downstream energy markets independently, without analysing the effects that developments in one market may have on the other. Our aim was to develop a holistic and integrated model that would expose and link the economics, physical interconnections and constraints across both of these market places.

The Concept
Recent independent studies indicate that gas demand is growing throughout Europe. However, it is not likely that the UK, Europe’s biggest gas market, can remain self-sufficient in gas. New analysis by Petroleum Services of Deloitte & Touche, reveals that the equivalent of another Ormen Lange – the largest new development offshore Norway (18 trillion ft ł) – needs to come on stream every year from 2007 just to meet existing UK demand (see Figure 1). The recent announcement by Marathon Oil to build Symphony, a 675km pipeline from Brae in the North Sea to Bacton in the UK, demonstrates the perceived opportunities for gas suppliers in the coming years.

Figure 1 – Future UK supply and demand

In order to give us a better understanding of the European energy market, both present and future, the gas toolbox was conceived. The specification was to develop a system to enable us to visualize the demand and supply characteristics of the European gas market and to economically model the gas network system. A number of scenarios would be created to model the effects of different market conditions. Under these market conditions, demand and supply figures could then be predicted over a twenty-year period. The GIS would allow users to visualise the changing supply and demand balance for the European gas market in granular detail. The specification included a network model of European gas transportation tariffs. This presented us with an opportunity to test the network analysis functionality of ESRI’s ArcGIS software.

The Approach
The requirements of the analytical and predictive components of the gas model were largely the driving forces behind the data and functionality that were chosen. From an early stage it was agreed that the economic models would be created within Microsoft Excel and the spatial data visualized in the ArcMap component of ArcGIS. An interface would need to be developed to allow the results of the spreadsheets to be uploaded into the GIS. The underlying Excel spreadsheets fell into three categories:

(I) Asset Based Foresting Model (ABFM)
A forecasting model for gas demand was developed in Microsoft Excel for each of the seven main gas markets in Europe (UK, Belgium, Netherlands, France, Germany, Spain and Italy). Each model contained up to four different scenarios of future gas demand, (i) High and flat gas prices, (ii) Low and volatile gas prices, (iii) A ‘green’ scenario that maximized the environmental contribution from gas, and (iv) a ‘secure’ scenario, where priority was given to the role of gas in long term energy supply security. A bottom-up approach was adopted that allowed us to focus on individual plants in each of these countries and allocate gas demand to gas reception terminals or import routes. At the same time, gas demand was integrated within the wider energy market context, taking into consideration coal, oil and renewable energy sources. The flexible nature of the ABFM enabled us to analyse development needs for future network and import/reception capacity investments as well as gas fired power stations in the electricity sector (see Figure 2).

Figure 2 - Calculating future gas demand for each power station in the ABFM

(II) European Gas Traiff Model (EGTM)
In addition to the ABFM, gas tariff models were created in Excel for each of the seven gas markets. The models were initially seeded with published tariffs for annual unmodulated gas delivered to city gate. However, they were designed to be flexible enough to accommodate changes, for instance if tariff systems altered at a later date, or if the user wanted to test their own assumptions about future tariff levels and regimes.

(III) Long Ran Breakeven Cost of Supplay Model The third Excel model detailed annual gas supply volumes and long run breakeven prices for discoveries, producing, planned and yet to find gas assets in the North Sea. Additional fields in Russia and Algeria were also modelled. Each supply asset was allocated to a gas reception facility on an economic basis. Sufficient supply volumes were identified and modelled to ensure that gas demand in the highest volume scenario would be exceeded by at least 10%.

The ‘Gas Tool Box’
During the development stages of the economic models it was clear that a great deal of thought would have to go into the way they interacted with ‘the toolbox’ (our customised ArcMap application). It was evident from the outset that each of the Excel models contained information that could be directly related to elements within the GIS. This assisted us in deciding which data layers should be included and the level of attribution assigned to each layer.

The following layers were added to the GIS based on relatable features in the Excel models:

Power stations:	Each power station was listed in the ABFM with an annual estimate ofgas demand over 25 years.
Terminals:	Gas reception terminals featured in the ABFM with annual estimates of gas supply. These figures in aggregate did not exceed volume estimates from the ‘long-run breakeven cost of supply model’.
Gas pipelines:	Each pipeline was classified by country and assigned a tariff value through the EGTM.
Entry/exit points:	Entry and exit points were also modelled in the EGTM for each country.
Oil/gas fields:	The supply characteristics and economics of each field were estimated using the supply model.

These layers, combined with cultural data, such as coastlines, urbanized areas and cities, formed the visual elements of the toolbox. Upstream data was sourced from our in-house upstream GIS system, PetroView®, whilst some downstream data was purchased from a third party vendor. Considerable time and effort was taken to integrate the two datasets, including spatial and attribute information. Once merged and cleaned up, the data was exported to a personal geodatabase (Access) and layer files created to specify the required styles and symbology. Further work was needed on the pipeline dataset and the interconnecting points such as terminals and power stations in order to create a gas network model.

The Gas Network Model
The European gas market infrastructure constitutes a network of pipelines and LNG shipping routes running between fields, terminals, LNG plants, storage sites, platforms and other facilities. The Excel models modelled each of these entities to a certain degree, but the real challenge was to develop a system that modelled the entire marketplace from wellhead to burner tip. We wanted to be able to calculate, break down and explore the costs involved in producing, shipping and delivering gas from any field to any point in Europe. With the underlying data stored in the supply and demand Excel models, our next step was to link this information into a network model in the GIS. The object-oriented data model in ArcGIS allows for the creation of geometric networks. A geometric network is a topological relationship between adjacent features. This lends itself well to a gas transportation network that can be built up of features that have connectivity relationships with other features around them. Using the ‘geometric network’ tool within ArcCatalog, a gas network model was built within the ‘toolbox’. Once built, the interconnecting nodes between each section of pipeline and LNG route were classified into one of six point features: (i) Entry points (terminals), (ii) Exit points (LDZ’s), (iii) Fields (oil & gas), (iv) Platforms, (v) Power stations (CCGT), & (vi) Pipe junctions. These were assigned names and country codes where possible. In a similar way, edge features were classified into gas, oil, condensate etc. pipelines, LNG routes, and proposed pipelines. The name, country, diameter and onshore/offshore status were then encoded. These core attributes now allowed us to join additional information to the tables from the Excel models.

Figure 3 - Visualising the infrastructure & predicting gas demand using the ‘gas toolbox’

Integrating the Models
The collection of software components that comprise ArcGIS is known as ArcObjects. More than 1,100 individual COM-based components are provided and customisation is performed using the built-in Microsoft Visual Basic for Applications (VBA) scripting capabilities or a COM-compliant programming language such as Visual Basic (VB), Visual C++, or Delphi. VB was selected as the programming language and by combining this with the ArcGIS data model, it was possible to develop routines within the toolbox that allowed us to interact with the Excel models. A series of ‘update’ style interfaces were developed that could load in demand and supply data from the spreadsheets, join these to the features in the GIS and populate their respective databases with new data. Initially, each feature could store up to 20 years worth of data. The GIS was set up to show all four ABFM scenarios at any one time, with each scenario constituting a separate ‘Data View’. This allowed users to view the ‘green’ scenario, for example, and then click on one of the other scenarios to explore the differences within a single session. Once the Excel models ere integrated additional functionality could be added.

Functionality
A number of basic tools were developed in VB to assist users with their enquiries. The first tool adds a temporal view to the ‘toolbox’. Users need to visualise the gas demand data and see how it changes over time. For instance, what would the gas demand picture look like in the UK in 2015? A list box populated with years was added to ArcMap, and a routine developed that refreshed the layers in the map when users selected a year. If a user clicks on 2015, then the map will redraw the demand layer using the values from the 2015 field from the database. Symbology is scaled according to the level of demand, so that the greater the demand for gas, the larger the symbol. In this way users can easily determine where gas demand might be greatest for any given region in Europe. Demand symbols are also classified by colour into one of three types: domestic demand, industrial demand and individual CCGT plant demand. This provides a snapshot for any given year of the gas demand market and enables users to explore the relative importance of individual power stations, and regional changes in the domestic and industrial markets over time. In a similar way, supply data is updated in the fields’ database, providing an opportunity to explore gas production estimates over time. Future development will allow users to select a gas field and view a production profile for it.

Search tools were deemed to be particularly useful such as ‘Find Field’, ‘Find Terminal’ and ‘Find Power Station’. Users can type a name into the Search dialog box and the map will zoom in to the feature. A series of pre-set views have also been added to make navigation of the map quicker and easier. It also assists users wanting to run the built-in network analysis function manually, although an automated system of doing this was also developed. The basic tools provide users unfamiliar with GIS systems, with a simple mechanism for exploring the data. The main value of the toolbox however, is derived from combining the gas network model in the GIS with the underlying Excel tariff models.

The Gas Rout Calculator
An interface has been developed that enabled users to explore the costs of producing, shipping or piping gas along a selected route across Europe. Users start by selecting a source gas field followed by a country to ship or pipe it to. Once the country has been chosen, they are presented with a list of reception terminals to bring the gas to shore (entry points), followed by a list of exit points, such as gas-fired power plants or Local Distribution Zone’s (LDZ’s). Depending on the country selected and its particular tariff structure, the interface requires further selections to be made, such as ‘Annual gas volume (Million mł)’ or ‘Floor Price (pence/kwh/day)’. Once completed, the model calculates the upstream and downstream components of the route. Different methods are employed to calculate these two components, as well as the downstream results that are also dependant on the tariff system of the selected country.

(I) The Upstream Component
Upstream results are calculated using the network analysis functionality of the GIS. Costs of production, offshore pipeline tariffs and intermediate processing and facility costs were added into the GIS when the gas network model was generated. Every pipeline and intermediate facility was assigned a weighting prior to building the network model. These weightings were actually tariffs derived from the ‘long-run breakeven cost of supply model’. When calculating upstream costs, the gas field and entry point are automatically flagged and a network analysis activated which searches for the lowest weighted route between these two points. The resulting selection represents the cheapest route between the source field and terminal. The upstream gas network in the toolbox is not constrained by the conditions that would typically be present in the real marketplace. The benefit of this is that users can explore alternative routes and options that may be available to them in the future, whilst allowing them to derive a ‘truer’ view of the actual value of a constraint. Alternatively, constraints can be added to keep the routing result more representative of the real world.

(II) The Downstream Component
Having calculated the cheapest route of transporting gas to shore, the model can then set about calculating the cheapest route from entry to exit point. It became apparent that there are essentially two types of tariff system in Europe. These systems are characterised by the UK market and the Belgian market. The UK marketplace is essentially dependent on point of entry and point of exit. A series of fixed prices have been set to determine the cost of transporting gas from an entry point to one of the many exit points across the UK. These exit points are either Local Distribution Zones (LDZ’s) or specific power stations. In this case, the gas model feeds the entry and exit points, and any of the other options, into the UK Excel tariff model. The resulting costs can then be retrieved from the Excel model without any need to interrogate the spatial data. The Belgian marketplace, however, works in a more complex way. The Belgian regime bases its tariffs on distance travelled and capacity of pipeline. In this case the calculations are more sophisticated. The entry and exit points are flagged on the map before activating a network analysis between the two points. A weighting is applied to the network analysis and the shortest path in kilometres selected. The selected route is subsequently split into HTL and RTL components based upon their classification in the geodatabase. The respective lengths are then fed into the underlying Excel tariff model along with the entry and exit points, and the downstream costs retrieved. Displaying and saving the Results
Once the model completes calculation of the total costs of the upstream and downstream components of the query, a further interface appears within the GIS showing a complete breakdown of the two costs. The upstream component is detailed in sequential order from the initial gas field down through to the reception terminal with costs in € per therm. The downstream costs are broken down into HTL and RTL costs, plus any additional costs such as fixed or term capacity charges. Downstream units are in € per gwh. The results of the query can then be saved as an Excel spreadsheet. An ‘Export’ button on the results interface dumps the results out at the same granular level of detail. Upon closing the interface, the user is then presented with a map of the result. This is in the form of a network analysis selection showing the complete route from start to finish.

Figure 4 - Snapshots from the gas tariff calculator (left) and gas network model (right).

The Benefits

Reviewing Costs
The gas route calculator enables users to assess what potential production, shipping / pipeline, and cross border costs might be incurred when sourcing gas from a particular field. Further development of the model will extend the analysis across multiple countries and gas networks. Users will be able to run a query through France and Switzerland to Italy, or from Algeria through Italy, Switzerland, and France and into the UK market. The temporal elements of the model allow the switching on and off of new or decommissioned assets respectively, depending on the selected year. The gas network model allows features to be enabled or disabled at runtime so pipelines, platforms or terminals can be excluded from or included in the network analysis. The effects of new assets coming on-stream can therefore be seen. For instance, how will the UK market be affected when Symphony comes on-stream in 2005 – what will the potential costs of supply be, and how will this compare with other UK import routes?

Planning New Sites
A fundamental consideration when populating the ABFM with data was that supply must always exceed gas demand by 10%, representing an upstream capacity margin. However, the ABFM exposed a mismatch of supply and demand at individual terminals and in some years, potential supply outstrips demand at some entry points. Assuming that all existing power stations were already running at full capacity, then if the laws of economics prevail, price pressure would occur to stimulate increased demand or shut-in excess supply. This would be flagged within the ABFM and users would be required to redress the balance, for instance, by adding a new power station into the Excel model. When the ABFM is imported into the GIS, users are faced with the question of where to build the new plant. The GIS then comes into it’s own because users must select a suitable site for their new plant. They might base this on surrounding demand centres, for instance, is there a region where demand is set to rise considerably and infrastructure is well established? How close to existing infrastructure (pipes, wires, storage sites, hubs), can I build my new plant, or are there any oil-fired power stations that have recently been decommissioned that I can convert to CCGT? Spatial queries such as buffering would allow users to search for new sites based on locational criteria, e.g. find me a UK site for my new power station within 20km of a major domestic demand centre, within 10km of the National Transmission System (NTS) and 10km of a wire to the National Grid. The strategic value of planning for future supply and demand scenarios using GIS in this way becomes apparent.

Conclusion
Existing gas infrastructure in Europe has been built over many years, generally by connecting a specific field to shore, rather than from a catchment area to the market. As energy market dynamics change and sources of gas in the North Sea mature, gas purchasers will start looking further afield. The ‘gas toolbox’ is an intelligent view onto this marketplace. The integrated nature of the model captures the entire gas value chain from wellhead to burner tip. Bringing together detailed Excel models with spatial data of the European energy market, the ‘toolbox’ provides a unique way of examining the complex relationships between gas demand and supply. The open and intuitive interfaces in the Excel and GIS models enables users to modify any part of the economic and spatial components thus avoiding a ‘black box’ approach. Temporal and spatial scenarios can be developed and explored over a 20-year period allowing users to value a range of risk and market outcomes. Constraints can be applied, or removed from the gas network model. Users can, therefore, explore new options in terms of alternative asset utilisation, new asset development, or model asset optionality over time. The integration of a gas route calculation system in the ‘toolbox’ adds a new dimension to the traditional network analysis functionality of GIS. This sophisticated tool combines network calculations, geodatabase attributes and Excel based tariff models. It can be utilised as an estimation tool using current published tariff systems or extended to be a forecasting tool with possible future tariff systems. The flexible and extensible nature of the ‘toolbox’ allows users to develop credible future scenarios of the European gas market, to frame gas market risk and to explore asset decision support options.