Fractal Applications: From Anaerobic Digestion To Urban Wind Turbulence Thomas Woolmington School of Electrical Engineering Systems Prof. Jonathan Blackledge Date 15/2/2013
Dec 15, 2015
Fractal Applications: From Anaerobic Digestion To Urban Wind Turbulence
Thomas WoolmingtonSchool of Electrical Engineering Systems
Prof. Jonathan Blackledge
Date 15/2/2013
In its simplest form Anaerobic Digestion is the decomposition of biological matter in an atmosphere that has no oxygen present.
This decomposition releases gases. For a good system circa 58% Methane, circa 40% Carbon Dioxide and traces of
Carbon Monoxide and Hydrogen Sulphide. These gases are produced as a by-product of bacterial interaction with organic matter. The bacteria thrive in stable PH environments circa 6.8-7.2 and at specific
temperature classes for different methanogen families
(NÍ Ruanaigh, 2011)
Anaerobic Digestion Basics
DEL Seminar Friday 15th February 2013
(Hamilton, 2012)
Digester Designs of Interest
DEL Seminar Friday 15th February 2013
Suspended Solid Up-flow Digester
Anaerobic Lagoon
(Wall et al, 2012)
(Worms, 2001)
Retrofit Current Slatted Units
DEL Seminar Friday 15th February 2013
Slatted Unit Construction Cross Section of Typical Slatted Unit
Finished Installation
Pros: Bovine agriculture in Ireland requires the use of storage facilities of rumen for a minimum of
16 Weeks (custom built tanks) Anaerobic Digestion has been proven to reduce the carbon equivalent footprint of
agriculture. Biogas can be used as a heat source on farms as the vast majority of farms in Ireland have
associated domestic dwellings. (If CHP is incorporated electricity can be generated also)
Cons: The Anaerobic Digestion process currently only monitors the environmental conditions of a
biological “Black Box”. As a result microbiologists manually monitor large AD plants. When CO2 equivalents are applied; agriculture is the largest greenhouse gas producer in
Ireland. Lagoon technology does not suit slatted unit tanks (currently used in Irish agriculture) or
indeed our Irish climate. The AD process can be simplified (passive) in warmer climates. Lack of awareness of the technology with Irish farmers.
Project Concept and Justification
DEL Seminar Friday 15th February 2013
Fractal Basics
Consider a simple trunk - branch - twig structure and add complexity of self symmetry along the structure
DEL Seminar Friday 15th February 2013
(Turner et al., 1998)
Fractal Basics
Note that the same shapes are present at different scales and orientation multiple times throughout the image
DEL Seminar Friday 15th February 2013
(Kulturservern, 2010)
Fractal Basics
• Self symmetry within structures can be quantified by means of fractal analysis (shapes within shapes).
• When we apply this concept to biological organisms it can in some instances be used as a means to quantify non-Euclidean (linear, area based or volumetric) growth.
• Consider leaves on a mature tree from season to season.
DEL Seminar Friday 15th February 2013
Fractal Basics
What would happen if the same concept was applied to Methanogen cell structures in active substrate within an anaerobic digester?
(Kubota et al., 2008)DEL Seminar Friday 15th February 2013
Correlation of simulation with
imageImage
Acquisition
Mathematical Simulator
Main Aims for the previous year!
2 strand approach; Develop a mathematical simulator of
fractal structures and determine its mathematical accuracy.
Acquire images of organisms that can potentially be used in the mathematical simulator.
Ideally the above two streams would merge in order to prove concept
Direct linking of organism growth rates to simulated model This gives a Quantitative value a reference to the biological growth state within a
reactor.This in turn can then be used to provide a closed loop means of monitoring the reactor process
Year 1 Aims and Methodology
DEL Seminar Friday 15th February 2013
DEL Seminar Friday 15th February 2013
Verification of concept of variable volume digesters
Cross between a lagoon and a suspended media up flow digester.
Submitted last June
(Images courtesy of Mr. A Bogle)
BSc Student Dissertation Supervision
Experimental Layout
Promising and potentially marketable
Needs a little mechanical alteration but the results speak for themselves
Any input would be greatly appreciated probably mechanical / manufacturing
(Image courtesy of Mr. A Bogle)
Problems Encountered
• We have a technology that does not appear to need monitoring!
• The Inability to obtain suitable images of organisms in situ.
• We have a maths model but nothing for it to be used on.
My PhD
DEL Seminar Friday 15th February 2013
Thought Process
• The Fractal model is deigned to look for self similarity within a noisy signal.
• Theoretically it should be able to quantify how noisy a signal is. Effectively more self similarity means more noise etc.
DEL Seminar Friday 15th February 2013
Enter Real world Data
• A chance discussion took place with Keith Sunderland
• Access to a large amount of 10Hz urban wind speed data.
• Add a little bit of tweaking to convert to a wind speed signal
• Put it into the current simulator and see what happens.
DEL Seminar Friday 15th February 2013
New Turbulence Metric
• Effectively the fractal model measures turbulence but it strictly speaking cannot be called fractal as a “fractal” in this instance should be bounded by a limit of Df=2.
• For comparison the “fractal” model was then compared to the standard turbulence metric (Turbulence intensity (TI)) for comparison.
• Then with the new metric longitudinal TI (effectively cosine corrected TI per 10 minute mean wind speed)
DEL Seminar Friday 15th February 2013
Unfiltered Data Comparison
• Unfiltered results 4502 samples• Note! TI>100% giving erratic spikes
DEL Seminar Friday 15th February 2013
Filtered Data Comparison TI>100% filtered out
• Filtered results 4327 samples• (3.88% of samples filtered out)• Note! Trending differences
DEL Seminar Friday 15th February 2013
Issues with the TI metric• An Investigation was made into the origin of the TI longitudinal metric and the
principles that it relies on in order to gain an understanding of the non correlation with the “fractal” model.
• Firstly TI is calculated based on,
• Secondly it was originally designed as a metric for large rural / offshore turbines where high wind speeds were present in low turbulence environments.
• It became obvious that the TI metric is unsuitable for use in an urban environment due to low wind speeds and high standard deviation making results appear to be asymptotically high in nature.
• But consider the following!
𝑻.𝑰.= 𝑺𝒕𝒂𝒏𝒅𝒂𝒓𝒅 𝑫𝒆𝒗𝒊𝒂𝒕𝒊𝒐𝒏 (𝟏𝟎 𝒎𝒊𝒏 𝒔𝒂𝒎𝒑𝒍𝒆)𝑴𝒆𝒂𝒏 𝒘𝒊𝒏𝒅 𝒔𝒑𝒆𝒆𝒅 (𝟏𝟎 𝒎𝒊𝒏 𝒔𝒂𝒎𝒑𝒍𝒆)
DEL Seminar Friday 15th February 2013
• Df =1 is effectively 0% turbulence
• However the current TI metric would classify this sample as having 31% turbulence
• The current TI metric does not allow for trends within the wind speed sample
DEL Seminar Friday 15th February 2013
Investigation into the TI metric• Having said that the TI metric has not been used to its maximum effect in terms of
power prediction.
• Previous studies have shown that for rural and off shore site that the PDF of wind speeds within a 10 minute sample follow a normal Gaussian pattern.
• Some of the experimental data shows that this is not the case in urban environments
logarithmic at low mean wind speeds
Rayleigh type for the majority of usable wind speeds
In a yet to be published paper co wrote by Keith Sunderland we have proposed that the following chart may be a generic case but as of yet we do not have enough site data to verify if this is true or not.
DEL Seminar Friday 15th February 2013
Investigation into the TI metric
• If this is the case all of the above PDFs can be encompassed by the Weibull distribution.
• The paper will outline how power can be accurately predicted for any turbine based on just a u mean and TI value for a 10 minute period. Thus dramatically reducing the amount of information that needs to be stored.
Mean uHigh Turbulence(urban/suburban)
Med Turbulence(Suburban/Rural)
Low Turbulence(Rural/Off-shore)
Low Exponential PDF Exponential / Rayleigh PDF Rayleigh PDF
Medium Exponential / Rayleigh PDF
Rayleigh PDF Rayleigh / Gaussian PDF
High Rayleigh PDF Rayleigh / Gaussian PDF Gaussian PDF
DEL Seminar Friday 15th February 2013
Some Interesting facts about the Weibull Distribution• Weibull analysis has long been associated as a means to predict component
failure rates in Mechanical and Manufacturing engineering.
• Up till recently the Weibull distribution model has been accepted to be correct based on numerous sets of empirical data.
• Weibull distributions have been shown to fit PDFs of numerous wind resource sites
• It has recently been proved by (Brown and Wohletz, 1995) to always have a Df that lies in the region of (0 ≤ Df ≤ 3).
Note! The last point now inherently links the statistical PDF model to fractal and levy index theory
DEL Seminar Friday 15th February 2013
• If this is the case all of the above PDFs can be encompassed by the Weibull distribution.
• The paper will outline how power can be accurately predicted for any turbine based on just a u mean and TI value for a 10 minute period. Thus dramatically reducing the amount of information that needs to be stored.
ReferencesBROWN, W. & WOHLETZ, K. (1995). Derivation of the Weibull distribution based on physical principles and
its connection to the Rossin-Rammler and lognormal distributions. Journal of Applied Physics, 78(4), 2758-2763.
HAMILTON, D. (2012). Types of Anaerobic Digestion [Online]. Extension.org. Available: http://www.extension.org/pages/30307/types-of-anaerobic-digesters#Contributors_To_This_Article [Accessed 1st Febuary 2013].
KUBOTA, K., IMACHI, H., KAWAKAMI, S., NAKAMURA, K., HARADA, H. & OHASHI, A. (2008). Evaluation of enzymatic cell treatments for application of CARD-FISH to methanogens. Journal of Microbiological Methods, 72(1), 54-59.
KULTURSERVERN. (2010). Modulations of Man [Online]. kulturservern. [Accessed 10th February 2013].NÍ RUANAIGH, A. (2011). Developing Anaerobic Digestion Cooperatives in Ireland. Master of Science
Energy Management, Dublin Institute of Technology.TURNER, M. J., BLACKLEDGE, J. M. & ANDREWS, P. R. (1998). Fractal Geometry in Digital Imaging, London,
Academic Press.WALL, G., HAMMOND, B., DONLON, P., JOHNSON, N. & SMITH, J. (2000). Commissioning and operation of
high rate anaerobic lagoon (HRAL) reactors. 63rd Annual Water Industry Engineers and Operators’ Conference. Warrnambool, Australia: Water Industry Operators of Australia.
WORMS, C. (2001). Concrete Slatted Floor Systems [Online]. Worms Lumber and Ready Mix. Available: http://www.wormsreadymix.com/slats.html#illus [Accessed 10th February 2013].
DEL Seminar Friday 15th February 2013
ReferencesBROWN, W. & WOHLETZ, K. (1995). Derivation of the Weibull distribution based on physical principles and
its connection to the Rossin-Rammler and lognormal distributions. Journal of Applied Physics, 78(4), 2758-2763.
HAMILTON, D. (2012). Types of Anaerobic Digestion [Online]. Extension.org. Available: http://www.extension.org/pages/30307/types-of-anaerobic-digesters#Contributors_To_This_Article [Accessed 1st Febuary 2013].
KUBOTA, K., IMACHI, H., KAWAKAMI, S., NAKAMURA, K., HARADA, H. & OHASHI, A. (2008). Evaluation of enzymatic cell treatments for application of CARD-FISH to methanogens. Journal of Microbiological Methods, 72(1), 54-59.
KULTURSERVERN. (2010). Modulations of Man [Online]. kulturservern. [Accessed 10th February 2013].NÍ RUANAIGH, A. (2011). Developing Anaerobic Digestion Cooperatives in Ireland. Master of Science
Energy Management, Dublin Institute of Technology.TURNER, M. J., BLACKLEDGE, J. M. & ANDREWS, P. R. (1998). Fractal Geometry in Digital Imaging, London,
Academic Press.WALL, G., HAMMOND, B., DONLON, P., JOHNSON, N. & SMITH, J. (2000). Commissioning and operation of
high rate anaerobic lagoon (HRAL) reactors. 63rd Annual Water Industry Engineers and Operators’ Conference. Warrnambool, Australia: Water Industry Operators of Australia.
WORMS, C. (2001). Concrete Slatted Floor Systems [Online]. Worms Lumber and Ready Mix. Available: http://www.wormsreadymix.com/slats.html#illus [Accessed 10th February 2013].
DEL Seminar Friday 15th February 2013
Any Questions?