Heriot-Watt University
ROS: The Research Output Service. Heriot-Watt University EdinburghNot a member yet
4689 research outputs found
Sort by
Detection of geobodies in 3D seismic using unsupervised machine learning
Get PDFIn this work, we present a novel, automated method for detecting geobodies in 3D seismic
reflection data, helping to reduce interpreter bias and speed up seismic interpretation.
A seismic geobody refers to a geometrical, structural, or stratigraphic feature, such as
a channel, turbidite fan, or igneous intrusion. Geobodies are subtle seismic features,
hard to pick, and their detection is challenging to automate due to their complex 3D
geomorphology and diversity of shapes. Nevertheless, the detection and delineation of
these structures are essential for improving the understanding of the subsurface as well
as building a variety of conceptual models.
In our approach, we can rapidly interpret large 3D seismic volumes using point cloud-based segmentation to identify geobodies of interest, including complex stratigraphic
features like lobes and channels. By converting the 3D seismic cube into a 3D seismic
point cloud (sparse cube), we reduce the volume of data to analyse, which in turn speeds
up the detection process. First, we build the 3D point clouds by filtering the seismic
reflection volume using different seismic attributes, and then each point in the cloud is
segmented into different clusters. The clustering is performed using the unsupervised
Density-Based Spatial Clustering of Applications with Noise (DBSCAN) which allows
the segmentation of all structures present into delineated objects. The clustered objects
can then be characterised by features based on their 3D shape and spatial amplitude
distribution. Finally, our method allows the selection of a specific geobody and can
retrieve geobodies based on their similarity to exploration targets of interest.
The method has been applied successfully to two modern 3D seismic datasets (Falkland
Basins) and two types of geobodies: fans and sill intrusions. We demonstrate that our
method can scan through a large 3D seismic volume and automatically retrieve likely
fan and sill geobodies in a very efficient manner. This approach can be used to scan
through large volumes of 3D seismic, looking for a wide variety of geobodiesJames Watt Scholarshi
Time-stepping methods for stochastic PDEs
Get PDFStochastic differential equations (SDEs) and stochastic partial differential equations (SPDEs) are fundamental modelling tools for a wide variety of disciplines
including finance, engineering and the physical sciences. Therefore, efficient and
accurate computational methods to approximately solve SDEs and SPDEs are of
great interest. It has been proven that standard time-stepping methods, such as the
Euler-Maruyama (EM) method, fail to converge strongly for SDEs or SPDEs which
contain a non-globally Lipschitz drift function. However the majority of applications
of interest to practitioners fall into this class, examples include the CIR model in
finance or the stochastic FitzHugh–Nagumo model in neuroscience.
In this thesis we introduce two types of explicit fully space-time discrete numerical methods to approximately solve semilinear SPDEs with non-globally Lipschitz
drift for both additive and multiplicative noise. To ensure stability, the first type
of method employs drift taming and the second utilizes adaptive time-stepping. We
prove bounded moments and strong convergence rates for both types of methods.
We use one of our numerical methods to construct a novel proof of existence for the
SPDE solution which includes rough multiplicative noise.
We test our numerical methods to the existing methods in the literature and
show significant improvements in both accuracy and computational efficiency for
several widely studied test equations
Elucidating the intracellular signalling pathways that mediate nanomaterial induced pro-inflammatory cytokine production in pulmonary cells in vitro
Get PDFThe aim of this study is to identify the optimal approach to study the intracellular
signalling pathways that regulate nanomaterial (NM) induced pulmonary pro-inflammatory responses in vitro. Submerged cultures of pulmonary epithelial cells and
macrophages and more physiologically relevant 3D air-liquid interface (ALI) models
were selected as these cell types are likely to interact with inhaled NMs.
The capacity of a panel of NMs to stimulate cytotoxicity, reactive oxygen species (ROS)
and cytokine production was assessed. Next, small molecule inhibitor (SMI) toxicity and
SMI inhibition of NM induced cytokine production was investigated. Furthermore,
immunocytochemistry, reporter cells, and western blotting methodologies were
compared for suitability of integration into tiered testing strategies. The physico-chemical
properties of NMs and their reactivity (ROS production) were assessed in parallel to the
hazard studies.
All NMs were negatively charged and agglomerated in relevant media. Sub-lethal
concentrations of NMs were identified from the cytotoxicity assays. Ag, CuO, and ZnO
NMs were shown to increase ROS production and increase pro-inflammatory cytokine
production (IL-8, MIP-2, TNF-α). The inhibitor of NF-κB kinase (IKK) inhibitors
reduced Ag and ZnO NM induced pro-inflammatory cytokine production in submerged
cultures but had no effect in ALI models suggesting that NF-κB activation downstream
of IKK may be important for NM induced cytokine production by pulmonary cells.
SMI/antioxidant inhibition of cytokine production was the most suitable of the methods
tested for integration into tiered testing strategies which assess NM hazard.James Watt Scholarshi
Bundling as a strategy for a commodity service brand introduction. The impact of bundle partner image on quality and risk perception and the role of complementarity
Get PDFIncreasing the quality perception and reducing the perceived risk of purchase improves
the chances of success for new service introductions. This research investigated whether,
for a new service brand introduction into the German residential electricity market,
bundling with a stronger service brand enhances the perceived quality and reduces the
perceived risk more than bundling with a weaker brand. In the goods category, it has been
scientifically shown prior to this research that bundling with a stronger brand achieves
this effect if the products are complementary. An academic knowledge gap in this area
existed because this enhancement effect was yet to be evaluated empirically for services.
This research applied price bundling to a new electricity service brand introduction via a
survey experiment with potential customers rating electricity bundle offers. The research
design was a 2*2 (brand image of bundle partner; complementarity) factorial design with
analysis of variance (ANOVA) to test the research hypotheses. The results narrow the
knowledge gap and contribute to professional practice by establishing that bundling with
a stronger brand enhances the perceived quality and reduces the perceived risk also for
services. The research furthermore demonstrated that complementarity is, independent of
the bundle partner brand image, a factor to improve quality perception and to reduce the
perceived risk of a new service brand
CO2 injection for enhanced gas condensate recovery : an experimental and theoretical study
Get PDFCO2 injection in gas condensate reservoirs has been identified as a viable technique
for alleviating condensate banking and enhancing gas condensate recovery (EGCR)
favoured due to prospects of storing CO2 in long terms. Multiple-contact miscibility
(MCM) with vaporising method is recommended to extract maximum gas condensate and
avoid leaving precious gas condensate fractions behind.
The research involves extensive PVT and core flood tests using CO2 and a binary gas
condensate fluid sample across a range of core permeabilities. Steady-state CO2-
condensate relative permeability data is gathered to improve H-n-P CO2 injection
simulations for enhanced gas condensate recovery and CO2 storage. The Schlumberger
E300 compositional simulator was employed to simulate incremental H-n-P CO2
injection, shut-in, and production cycles, mirroring laboratory experiments. The
simulation results emphasize the importance of using accurate CO2-GC kr data and
accounting for compositional changes during H-n-P CO2 injection.
In the following chapters, a practical framework was suggested based on the results to
accurately identify and quantify the effects of CO2-GC interaction during CO2 injection
for enhanced gas condensate recovery. Furthermore, the miscibility pressure of CO2 and
gas condensate sample was optimised to enhance the swelling and vaporisation
mechanism and determine the best injection scenarios at pressures below and above the
dew point for optimal gas condensate recovery and CO2 storage purposes. The
hydrocarbon recovery efficiency of the suggested injection technique for EGCR was
tested on high to ultra-low permeability core samples. The recovery efficiency of this
optimised plan was observed to surpass the conventional H-n-P CO2 injection albeit with
five times less the volume of CO2 required during the conventional injection approach.
The volume of injected CO2 was constrained by pressure limits over which the variation
in maximum condensate saturation is minimal. Results indicate that condensate recovery
significantly improves, reaching 69.7% after the fourth H-n-P CO2 injection cycle, with
49.5% additional condensate recovery post primary depletion phase and 48.6%
cumulative CO2 storage. At the end of the proposed H-n-P CO2 injection, the total gas
produced had an 85.9% and a 14.1% hydrocarbon and CO2 content respectively.
The experimental data reported in this thesis allow bridging the gap between
conflicting reports on the CO2-GC fluid interactions at pressures below and above the
dew point pressure (PDew) and provides a solid cornerstone to design optimised H-n-P
CO2 injection scenarios for EGCR and CO2 storage purposes
An open access carbonate reservoir model
Get PDFThis thesis presents a new open-source carbonate reservoir benchmarking case study, the
COSTA model, that uniquely considers major uncertainties inherent to carbonate
reservoirs, providing a far more challenging and realistic benchmarking test for a range
of geo-energy applications. The COSTA model is large (1,600 km2
), with many wells
(447) and large associated hydrocarbon volumes (109
bbls). The dataset embeds many
interacting geological and petrophysical uncertainties in an ensemble of model concepts
with realistic geological and model complexity levels and varying production profiles.
The large number of models and long-run times creates a harder computational challenge
than older benchmarking models.
The COSTA model takes inspiration from the shelf-to-basin geological setting of the
Upper Kharaib Member (Early Cretaceous), one of the most prolific aggradational
parasequence carbonate formations sets in the world. The dataset to build the model uses
43 wells from fully anonymized published data from the north-eastern part of the Rub Al
Khali basin, a sub-basin of the wider Arabian Basin. My model encapsulates both the
large-scale geological setting and reservoir heterogeneities found across the shelf-tobasin profile (~36,000 km2
), into one single model (~8,300 km2
), for geological
modelling and reservoir simulation studies.
The result of this research is a semi-synthetic but geologically realistic suite of carbonate
reservoir models that capture a wide range of geological, petrophysical, and geomodelling
uncertainties and that can be history-matched against an undisclosed, synthetic truth case.
The models and dataset are made available as open-source to analyse several issues
related to testing new numerical algorithms for reservoir characterisation, uncertainty
quantification, reservoir simulation, history matching, robust optimisation, and machine
learning. The novelty of my work is the provision of a unique and realistic open-access
dataset that enables reproducible science in the field of reservoir characterisation and
simulation and offers new training opportunities in the areas of reservoir characterisation
simulation and prediction of the reservoir performance of carbonate reservoirs. The
model(s) can also be used to study geological sequestration of CO2, the feasibility of
EOR processes, geothermal and groundwater studies
Situated grounding and understanding of structured low-resource expert data
Get PDFConversational agents are becoming more widespread, varying from social to goaloriented to multi-modal dialogue systems. However, for systems with both visual
and spatial requirements, such as situated robot planning, developing accurate goaloriented dialogue systems can be extremely challenging, especially in dynamic environments, such as underwater or first responders. Furthermore, training data-driven
algorithms in these domains is challenging due to the esoteric nature of the interaction, which requires expert input. We derive solutions for creating a collaborative
multi-modal conversational agent for setting high-level mission goals. We experiment with state-of-the-art deep learning models and techniques and create a new
data-driven method (MAPERT) that is capable of processing language instructions
by grounding the necessary elements using various types of input data (vision from
a map, text and other metadata). The results show that, depending on the task,
the accuracy of data-driven systems can vary dramatically depending on the type
of metadata and the attention mechanisms that are used. Finally, we are dealing
with low-resource expert data and this inspired the use of the Continual Learning
and Human In The Loop methodology with encouraging results
Initiating professionalisation in family firms through the identification, assessment and development of competency models : evidence from Ghana
Get PDFThe literature suggests that the professionalisation of the management of family owned firms
can be the best option to initiate family firm professionalisation and that this can be achieved
through competency modelling and assessment. For this study, the core research question
was: What are the challenges encountered in using competency frameworks to initiate the
professionalisation process of family firms in an emerging economy? To answer this question
the process of competency modelling and assessment was investigated through a positivist
multiple-case study design in both external cases and the case organisation. The case
organisation here is a family-owned business in its second generation ownership. The data
for the study were obtained from 48 completed competency ranking questionnaires and 14
completed 360-degree feedback assessments from six external organisations. For the case
organisation, the data were obtained from 14 participants. The study found that highly
structured and intensive competency approaches can be successful, but not without
challenges that were encountered. These challenges, if not addressed, may limit the potential
of the approaches. Recommendations include the adequate sensitisation of all stakeholders,
the adoption of company-specific competencies, the inhouse definition of behavioural
anchors and the institution of a shared vision and values. Further study could investigate the
attitudes to such structured and intensive competency approaches among stakeholders, the
application of this process in small organisations and the effects of customisation and
alignment of competencies with the vision and values of the organisations. Finally, more
guidance on how case study research can be carried out by owner-managers in their own
organisations would be beneficial
Improved practices for client contact management in a crisis during restrictions on personal contact as caused by Covid-19
Get PDFThe purpose of this thesis is to explore, over the course of the current pandemic and the
corresponding contact restrictions, changes to client communication practices in
professional services and the lessons that can be learned from the process. The overall
aim is to explore how business consultancies have developed crisis responses around
client communication, what their experience with these measures were and on the basis
of this, to evaluate which could be used in order improve crisis management and
preparation in the future or improve practice in general.
While much has been said about communication about a crisis, current research lacks
information on crisis management around client communication, especially in the
business sector the thesis investigates and with a focus on interactive communication
modes. There is also very little material on what is likely to become improved practices.
The present thesis seeks to address this gap.
Adopting a Critical Realist paradigm and an abductive research logic, a case study on the
consulting industry was conducted, with qualitative interviews as the main source of
primary data and thematic analysis as the method. Interviewees were professionals from
different types of consultancies from Germany, Austria and Switzerland. The industry
was chosen because for it, regular client contact and continuous close coordination are
key, and this exchange has been done in person in the vast majority of instances.
Key outcomes are a framework of recommendations for client communication in a crisis
and first insights into how current crisis management measures might affect the
workplace design after the crisis ends. Recommendations highlight in particular the
importance of infrastructure and preparation, the need for rapid and pragmatic decision
making, the usefulness of a project by project approach where possible, the central role
of stakeholder involvement and the potential need of also supporting clients.
This contributes to theory by presenting a more comprehensive framework on
communication with clients in a crisis than hitherto published. It contributes to practice
by providing a set of implementable recommendations for upholding and managing said
communication. This might also be applicable for crisis communication with other
stakeholders than clients and beyond the consulting industry
Contribution of tidal energy to an integrated island energy system
Get PDFTidal stream energy is an emerging sector of the energy generation industry. Compared to many other renewable energy resources it has high potential to provide base-load
power due to the predictability of the speed and direction of tidal currents. However, the
practical use of tidal stream energy requires extraction that is both efficient and appropriate; in engineering design, social impact; and economic viability. This study designs a
tidal array for Orkney waters, testing it against a wide range of constraints from engineer ing efficiency to market suitability. It explores various approaches to achieving efficient
energy extraction. Different energy generation patterns are examined to find the strategy
that best fits the pattern of energy demand of the islands, without conflicting with the
existing supply. The study demonstrates the potential of integrating tidal energy into an
island energy system without the need for expensive grid upgrades. It shows that arranging the turbines in a staggered sub-array (SSA) layout, and regulating the power output
of the tidal device, increases the capacity factor of the installed system. This strategy
improves the economic viability and commercial competitiveness of tidal energy.James Watt scholarshi
Access Repository Dashboard
Do you manage ROS: The Research Output Service. Heriot-Watt University Edinburgh ? Access insider analytics, issue reports and manage access to outputs from your repository in the CORE Repository Dashboard!