Bangor, Gwynedd a Humming Paradox OR ‘If the LFN phenomenon the world over known as the Hum is associated with renewable energy, especially wind power, why in Bangor does it appear to maximise when total UK wind energy output troughs but does not cease?’ By Dr Chris Barnes, manager Bangor Scientific and Educational Consultants e-mail manager@bsec-wales.co.uk First published without full reference list February 2016.

Abstract

The question ‘if the LFN phenomenon the world over known as the Hum is associated with renewable energy, especially wind power, why in Bangor does it appear to maximise when total UK wind energy output troughs but does not cease?’ is posed and theoretically answered. Experimental data based on personal observation of the Hum and real-time behaviour of wind generation and the National Grid generally support the answers. Once again an undeniable connection between the Bangor Hum and Power Systems has been demonstrated. A observation that the Hum maximises at lulls in wind energy but not with total cessation of wind generation has been experimentally supported. The connection between the Bangor Hum and Dinorwig Pumped storage scheme has also been supported by this work. A new observation is the connection between the Hum and power flows in HVDC interconnectors has been established and tentatively explained. Since we now have a ‘Smart Grid’ it ought to be possible to devise power flow algorithms to either a) reduce or eliminate the Hum completely or b) distribute its e3ffects so no one part of the country is overburdened by them.

Introduction

The Hum is an unusual LFN phenomena first reported extensively in Britain in the 1970’s and later in the USA in the 1990’s. Although some properties of the Hum, in particular descriptions of the perceived auditory effects such as for example idling engine noises and pulsating buzzing by those afflicted appear at first sight to be those of Infrasound (refs), a distinction has been drawn by Leventhal in the Hums are described as LFN where the source cannot readily be traced.

The world temporal evolution of the Hum can only be conveniently explained by a complex hypothesis involving interaction of the World’s power systems. Indeed, the Hum is now reported in virtually all corners of the globe and only countries which do not use significant amounts of renewable energy and do not have PME mains electricity earth systems appear to have escaped (refs).

To fully account for all the propagation and perception properties of the Hum including actual first hand and anecdotal reports one reaches the conclusion that the Hum is far more than just a noise and may, in fact, be augmented by a magnetic and or gravitational means of perception in addition to just normal audition (refs).

Based on visits to a website which is a database for logging instances of the Hum around the world, it would seem that the phenomenon is presently most associated with renewable energy, particularly the use of wind energy and in some cases pumped storage hydro-power, particularly employing Francis turbines ( ref Bangor and Italy). These types of energy are more likely to be able to supply the complex sets of coherent or quasi-coherent acoustic, infrasound, seismic and electromagnetic or electro-gravitational signals which constitute the Hum (refs). For example, I have previously shown that some high voltage power grids exhibit instability and may even act as electro-seismic parametric amplifier systems under excitation by wind energy and when driving or being driven by Francis Turbines ( refs).

Other types of signal may also complicate the Hum, for example there may be present acoustic, infrasound or seismic signals arising from other public utility services ( refs) and/or the Hum may have a preternatural component ( ref).

The presence of such multiple sources contributing to the Hum is thought to be the case in Bangor and has been suggested as accounting for its complex behaviour as a function of wind speed and direction (ref). There is even the possibility of Bragg matching with certain radio frequency sources (ref).

The paradox here in Bangor and elsewhere appears to be that the Hum maximises during lulls of overall UK wind generation capacity yet where wind generation output does not cease completely. The real-time data to show this has only recently become available on the internet ( REF).

It is the purpose of this brief publication to attempt to elucidate this paradox and to explore if there may be any alternative explanation to the complex behaviour of the Bangor Hum as a function of wind speed.

It is convenient to consider each of the potential Hum component sources in turn and as if arriving from wind energy. One must then consider the effect of local and overall wind speed on the generation and propagation of such components.

Acoustic sound and Infrasound Windfarms ( propagation)

Wind farms are known sources of acoustic sound and infrasound. The acoustic sound is known to be more annoying when wind farms exhibit OAM ( other amplitude modulation). The peak part of the OAM spectrum is …. Hz and would not be expected to propagate very far. There are, however, LFN and infrasound components to some OAM and wind farm emissions which can travel significant distances of tens of kilometres at night in a stable boundary layer.

A stable boundary layer will be destroyed by excessive turbulence, i.e. excessive wind speed and hence if windfarm emissions are a component of the Hum they would not be expected to propagate far in very strong winds. The wind farms which could influence a Hum in Bangor are at …… and North Hoyle , some …. Km and …..Km distance.

Thus propagation of sound from a windfarm is favoured by no or gentle wind rather than strong wind. Clearly if there is absolutely no wind, no emissions will be generated.

This behaviour supports the experimental observation that Hum in Bangor peaks during Lulls in total wind power generation.

Windfarms (generation)

When wind speed fluctuates or when wind is variable in direction windfarms are more likely to generate OAM and infrasound. Such fluctuations in speed and sudden changes in direction which cause more OAM and Infrasound will be accompanied by Lulls in total wind power generation and this too supports the observation.

Power grid Electromagnetic emissions ( Harmonic, sub and inter-harmonic )

Lulls in total wind power generation must logically be accompanied by wind speed reduction at some but not all wind farms. When the wind speed reduces and changes direction suddenly not only is OAM generated but power disturbances are also maximised (refs). Phase imbalance will be maximised due to way wind farms are laid out. Ground current flows are also maximised. Harmonic emission to space will also be maximised.

If magnetic perception is relevant or partially relevant to the Bangor Hum it too should maximise during such ‘Lulls’.

Seismic signals

The main seismic source in Bangor which could contribute to the Hum is Dinorwig. During Lulls in wind power generation at night Dinorwig represents a large load at the end of a long transmission line. Electrical and hyrdo-acoustic instability will tend to go hand in hand and any transmitted seismic signals will reflect this.

Experimental

The author keeps a careful log of Hum intensity and type ( i.e. the extent to which it appears to incorporate magnetic component) and has done for several years. The most profound observation is that ‘magnetic ‘ Hum outbreaks which were once very rare now appear to be much more commonplace especially during January 2016. It is noted that reactor 2 at Wylfa ceased generating on 25 April 2012 and reactor 1 stopped generating on 30^th December 2015.

Based on all my prevoius hypotheses of the Hum I would expect the Hum to get worse after these events as the power grid in this area will be significantly weaker particularly at night when Dinorwig is pumping.

It has only become possible to make a more precise estimation of the effect of wind generation on the Bangor Hum since the advent of the availability of real-time data at

http://winderful.diascreative.net/ an example is shown below.

And at :

http://nationalgrid.stephenmorley.org/ where also a more complex breakdown of all types of UK electricity generation can be found togather with details of interconnector imports/exports.

Using the above site data has been obtained for the month of January 2016 and compared with the subjective Hum level at the author’s home. It should be noted that there was never a day when there was no wind generation and wind generation at any one time varied from 0.14 -5 GW across the period.

At first sight it would seem that by extrapolation Hum level maximises at zero wind generation but since there was never a time when zero generation took place it would be unreliable to deduce the same. However, it is safe to assume that Hum levels peak at the minimum value of generation and that result is generally in support of the proposed hypothesis. Furthermore the trend for decreasing Hum levels with increasing wind generation could be synonymous with increasing wind speed and hence increasing masking noise.

I have previously suggested that the Bangor Hum is comprised of multiple Infrasound and LFN sources and have shown a complex behaviour ( sinusoidal) for each source as a function of wind speed (ref). If the sources are wind farms or influenced by the acoustic or electrical output of windfarms I would expect a polynomial solution to the above function ( subjective Hum as a function of wind generation). I propose a minimum of say a cubic function on the baits of multiple propagation paths and maybe this could be scaled to a 6^th degree polynomial the basis of differences of wind speed at the two most local wind farms.

P Value Results

r=.53 DF=29

The two-tailed P value equals 0.0022

By conventional criteria, this difference is considered to be very statistically significant.

Applying this polynomial suggest that there is a finite but weaker Hum level in the absence of wind generation. High wind speed can presumably provide masking noise and disrupt the coherence of Hum from whatever cause.

As wind generation falls, other methods of generation are brought on line. The plot below shows how conventional hydro- power ( not pumped storage ) can influence subjective Hum levels in Bangor.

The website http://nationalgrid.stephenmorley.org/ does not give detail of power used whilst pumped storage is pumping but merely tells how much is generated on a given day.

An assumption previously made in Bangor and Italy ( and with substantial experimental evidence together with theoretical support) is that Francis style turbines lead to Hum.

Thus if the residual Hum not due to wind power is due to Pumped Storage one might expect a good correlation between a night’s subjective Hum and the next day’s use of pumped storage energy since predictive algorithms are commonplace in the power industry.

P =.05 i.e. just statistically significant.

P=.038 i.e. more statistically significant.

D.C. Interconnectors

D.C. Interconnectors were first introduced to the UK Electricity Grid in the 1960’s. A 160 MW cross channel system (1961) was first introduced, superseded by a 2 GW system in 1986.

The HVDC Moyle Interconnector is the HVDC link between Auchencrosh, South Ayrshire in Scotland and Ballycronan More, County Antrim in Northern Ireland, which went into service in 2001. It is owned and operated by Mutual Energy.

The HVDC Moyle has capacity of 500 MW. It consists of two monopolar 250 kV DC cables with a transmission capacity of 250 MW each. The converter stations are completely equipped with light triggered thyristors. As of October 2011, the interconnector was out of service, repair efforts were and the cable became operational again with 450 MW in February 2012

BritNed Development Limited is a joint venture of Dutch TSO TenneT and British National Grid and operates the electricity link between Great Britain and The Netherlands which was commissioned in 2011. The maximum power rating of the BritNed interconnector is 1GW.

The 500 MW East West interconnector between Wales and Ireland was completed on 20 September 2012 it was inaugurated in Meath by UK secretary for energy and climate change Ed Davey, Irish prime minister Enda Kenny and European Commissioner for Energy Günther Oettinger. It is meant to allow better utilisation of energy, particularly renewables.

Such interconnectors can cause excessive and abnormal harmonics, see for example Ainsworth (1967) and thus could be implicated in the Hum. See also Larsen et al (1989). Although grid interconnection can reduce voltage flickers it is notorious for other types of degradation of power quality, see Shilpi et al (2013). Non-characteristic frequencies such as inter-harmonics can also occur with HVDC systems, see Hulme et al (2003). Inter-harmonics can provide unwanted acoustic noise in power systems, see http://admin.copperalliance.eu/docs/librariesprovider5/power-quality-and-utilisation-guide/311-interharmonics.pdf?sfvrsn=4&sfvrsn=4. See also Lin (2011). I have also commented extensively on this elsewhere, http://www.drchrisbarnes.co.uk/POWERGRID.htm.

Although theoretically HVDC Interconnectors ought to damp inter-area oscillations, few practical examples exist, see Vural (2016).

The site at http://nationalgrid.stephenmorley.org/ contains details of imports/exports on the various UK Electricity interconnectors. I have explored the relationship between these values and the relative Hum intensity during January 2016.

Above: Power Flow Irish/Wales DC Interconnector Circa 9 Hum level points range. R=.6

P Value Results

r=.6 DF=22

The two-tailed P value equals 0.0019 so by conventional criteria, this difference is considered to be very statistically significant.

Above: Cross Channel Power Flow Circa 1.5 Hum Level Points Range. R=.42

P Value Results

r=.42 DF=15

The two-tailed P value equals 0.0933

By conventional criteria, this difference is considered to be not quite statistically significant.

Above Britnet Power Flow Circa 8 Hum Level Points Range.

P Value Results

r=.34 DF=15

The two-tailed P value equals 0.1818

By conventional criteria, this difference is considered to be not statistically significant.

Above: Moyle Interconnector Power Flow Circa 4.5 Hum Power Points Range. R=.27

P Value Results

r=.27 DF=20

The two-tailed P value equals 0.2243

By conventional criteria, this difference is considered to be not statistically significant.

Interconnectors Discussion

In each case a quadratic equation provides best correlation. Perhaps to be expected given the potential for two way power flow. Although only the Irish Interconnector produces a statistically significant result, the result from the Cross Channel Connector is almost significant. The large scatter is due to the fact that multiple power sources in each network all contribute separately and singly in some way to Hum behaviour and there will be no too days in which airborne acoustic and ground borne seismic and space-weather are identical.

The results are very revealing and support all my prevoius work and deductions on the Hum. It is perhaps not surprising that given the proximity of Bangor to both the power grid and the sources of renewable energy that feed the Irish Interconnector that it should have most effect on the Bangor Hum. Indeed the Hum minimises when there is no power flow in that Interconnector. It is interesting that Hum maximises when there is maximum import or export of electricity across the Interconnector. Harmonics and inter-harmonics and radiation to space, ground currents and acoustic noise will all potential facets of the Hum will maximise and power quality will minimise under these conditions and almost uniquely with the Irish inter-connector wind energy will be one of the major sources of power from either side of the Irish Sea. I have previously raised the issue of sporadic Hums cropping up in various parts of Britain after changes in our National Grid system which will affect power flows, such as closures of certain power stations( ref). This present work certainly seems to reinforce those ideas.

Although not too statistically relevant, Hum level seems to minimise when Power Flow from the cross-channel Interconnector maximises. Presumably the stronger European Grid confers additional stability onto the UK grid or alternatively and additionally we are generating less renewable energy at these times. A similar although not statistically relevant effect is noted with the Britned Interconnector. The result with the Moyle Interconnector is also not statistically relevant but suggests that Hum levels in Bangor minimise at Maximum export to Northern Ireland. Two reasons could explain this. Firstly less dirty electricity is flowing into the National Grid from Scotland at these times and/or secondly very high wind levels are allowing such export and when the wind blows hard in Scotland, it usually blows form the West or South West and will be blowing similarly in North Wales and providing masking noise and disturbing coherence of Hum signals.

Conclusions

Once again an undeniable connection between the Bangor Hum and Power Systems has been demonstrated. A observation that the Hum maximises at lulls in wind energy but not with total cessation of wind generation has been experimentally supported. The connection between the Bangor Hum and Dinorwig Pumped storage scheme has also been supported by this work. A new observation is the connection between the Hum and power flows in HVDC interconnectors has been established and tentatively explained. Since we now have a ‘Smart Grid’ it ought to be possible to devise power flow algorithms to either a) reduce or eliminate the Hum completely or b) distribute its e3ffects so no one part of the country is overburdened by them.