The Hum: models for generation of and perception of an  LFN noise nuisance with intriguing properties, a  facet of renewable energy systems and their connection with the power grid at both local and more distant points , by Dr Chris Barnes Manager at Bangor Scientific and Educational Consultants   E-mail

  First published on-line Tuesday 15th January  2013 





The Hum, its LFN like nature, history and supposed origin in the Worlds power systems and effects are briefly reviewed.  Frequency components in the Worlds’ power systems are shown as adequate to provide those described for the Hum.  The wasted power in the World’s power systems is more than adequate to provide the Hum.  It strongly modulates the elctrojet.  Mechanisms of Hum generation are discussed and it is shown how the Hum can break out either close to a source of renewable energy or more distant from it.  A simplified model of Hum reception by buildings is discussed and perception therefrom by hearing and even deaf persons.    Finally possibilities for mitigation are discussed.




The Hum is a geo-sporadic noise nuisance heard/perceived by an estimated 2% of the world’s population, this figure probably rises to well over 10% in the 40-70 age groups.  The noise, an irregular pulsation like very low pitched buzz, can be extremely irritating and annoying and in its extreme even cause stress related illness and sleep deprivation. It is   mainly but not exclusively in buildings  at night even with the power switched off and  is characterised by the sound of a slowly and irregularly (quasi –periodically) idling engine and these days is perceived by the present author as often more sharply pulsing than musical.  Those musically minded hearers of the first instance of the Hum in the USA, the so called Taos Hum tone matched it between 30-80Hz and stated that it has quasi-period modulations or fluctuations of between 0.5 and 5Hz.  Others describe a higher pitched buzzing yet still with the same sort of underlying modulation.      Another strange facet of the Hum is that is either difficult or impossible to audio record.  Sadly this has pushed the Hum, for some, into the realms of Science Fantasy. This paper, as with the other works of the present author, aims to demystify the phenomenon. 


Everything but power systems has officially been blamed for the Hum. Scientists at Southampton University concluded it is simply distant noise.  Wikipedia blames amongst other things colliding waves under the ocean.  Applying critical thinking one immediately wonders why then distance noise or ocean waves shouldn’t cause the Hum in the bulk of Russia and Africa and other places which either don’t or until recently haven’t experienced the Hum.  


Contrasting all kinds of infrastructure between places that don’t and do experience the Hum and using world mapping it has recently been shown that the Hum is best correlated with sites wherein there is connection of renewable energy to the power grid.    Long time amateur Hum investigator John Dawes has suggested power systems may be to blame since the first Bristol and Largs Hums in the UK in the late 1970’s and early 1980’s but has never been able to be specific about which elements of such systems were the cause, instead John has referred to the general upgrading of the grid which took place in that era.   John does however have quite a controversial theory of how the Hum might be perceived in terms of gravity modulation.  The present author can see at least three physical mechanisms as to how this can occur but not everyone might understand or agree.   It is imagined that John came up with his hypothesis to account for  lack of audio recordings and to explain  how some deaf people claim to hear the Hum (and presumably describe the same noise as hearing people) and as to accounting for how the Hum is sometimes difficult to screen with ear plugs.


The present author feels we now should re-examine models of Hum to see if any simpler models can apply to perception in hearing people or if additional facets are needed for or enhance perception generally.


Hum generation

First we must ask is it feasible to accept the notion that the power grid with connected renewable energy sources is the cause of the Hum?  

Two questions need to be asked. Firstly is the power grid a big enough source of energy?  Secondly does the power grid provide the correct frequency elements?  


The world’s energy grids now carry power in the order of several TW of which about 200 GW is wind power.   Loses are reckoned to be about 7% of which known loses amount to 6.6%. This leaves 0.4% or   about 20GW.  World   power grids are cable of radiating electromagnetic energy, airborne acoustic energy and ground-borne seismic vibrations presumably this 20GW is distributed between the three.


It is interesting to compare the energy in the grid with lightning. It is estimated that a moderate thunderstorm produces about 500MW or power. There are about 2000 thunderstorms on earth at any one time =1TW.  Thus the world’s power grids exceed this figure.   


Thunderstorms influence space weather by producing magnetospheric whistlers which can be received as VLF signals on earth.  Power line harmonics and even ripple signals injected into power lines can also influence the magnetosphere and be received anywhere on earth.


The author thus feels it is fair to conclude that the power   grid is a large source of energy to provide a Hum or Hums over wide areas.   Consistent with this is that some outbreaks of Hum occur at quite a distance from power lines or electricity distribution equipment.   Kero at al (2004) have stated that in the vicinity of power lines, their harmonic signals are decades above those of natural signals due to global lightning discharges.       Even the magnetic component of the weak ripple control signal at 1050 Hz Dunedin New Zealand can be detected by the Demeter Satellite.


Dealing with the second question, it can be shown that power grids contain exactly the right range of frequencies for the Hum and sharply pulsating behaviours.  Firstly they contain a frequency of 50 or 60 Hz and its entire harmonics in various proportions.  These can be radiated in all or any of the forms detailed above.  Other frequencies can be supplied as a result of inter-harmonics, sub –harmonics and injected control frequencies such as those due to ripple control and cyclo-control.   


The Hum is described as having quasi-periodic modulations in the range 0.5-5Hz. Power systems can now supply the same which will appear manifest as frequency and voltage fluctuations.   Prior to the mid 1970’s power systems oscillations were not recorded in Britain, neither was the Hum.      At about the advent of the installation of pumped storage schemes, power systems oscillations of about .5 Hz were noted across the entire UK network.   Intra plant oscillations can also occur at 2-3 Hz. Local mode generator frequency swings occur at 1-2 Hz. Inter area oscillations occur at 1 Hz or less and finally torsion mode generator shaft oscillations can occur at anywhere between 10-46 Hz.   Complicating the situation for people living near Pumped Storage plants there will also be hydro-dynamic standing pressure oscillations between 2-3 Hz, these cause both hydro-acoustic pressure and initiate power oscillations (Koutnik et al 2006).   For wind turbines there is also infrasound at blade crossing frequencies and coherent voltage flicker effects at the same frequency.  


Finally the Hum has become ‘pulsier’ in recent years.  Power systems harmonics have also become ‘pulsier’ due to the use of more solid state switching, frequency and voltage compensation, AC/DC converters, Solar PV and wind-turbine inverters. On the super grid SVC’S dominate for creating harmonics but on the lower voltage networks embedded generation and domestic and commercial loads dominate (Lundquist –PhD Thesis Chalmers University of Technology, Sweden).     In the specific case of Solar PV converters, the panels themselves also have the undesirable capacity to act as radio antennas thereby picking up additional unwanted pulses from mobile and other communication networks, which can on occasion appear in the inverter output waveform.   


Another potential complicating factor for the Hum is the way in which the National Grid is now handling dynamic load response. Domestic refrigerators (since 2009) and Supermarket heating and ventilation equipment (since 2011) are now being used to this effect.  Since effectively they are a reactive load dynamic in dynamic response this too will increase the harmonics present.


Many loads in Britain are switched by the BBC Radio 4 Long wave Radio teleswitch. This doesn’t just send out a message to change Tariffs but is constantly signalling devices thorough the electricity network.  


A ripple control system called Pulsadis has been operated by EDF Energy in France for some 30 years and a similar system called VDEW operates in Germany.  These signals start late at night and persist until morning for Tarrif control. Similar systems operate throughout Europe and Australasia.    Signals from these can be received together with mains harmonics by radiation and  ionosphere reflection over huge distances.   It is not known is Pulsadis etc. features in the Hum or in its intensification but parametric mixing is always a possibility.      


Inter-harmonics are as relevant to the Hum as harmonics and sub-harmonics and may be caused electric motors, wind turbines, static frequency converters and mains-mark systems (Hanzelka and Bien 2004). Hanzelka   and Bien also describe inter-harmonics specifically as ‘a source of acoustic noise’.   


Hum perception – hearing person with building ‘amplification’


The present author has previously shown that different acoustic frequencies exist inside and outside buildings and on different walls and windows of buildings.  Buildings amplify background sound by room resonance.  The notion is that any Hum arriving may arrive by more than one path. In the simplest sense, airborne sound and seismic vibration can either arrive directly from any piece(s) of hardware in a power system carrying the relevant frequencies,   various combinations of which have been previously shown by the present author to elicit the Hum especially if the 50 Hz component is either pulsating or is not constant in frequency or amplitude (refs) .   The seismic component will set the building walls into forced or resonant vibration.  


An alternative hypothesis is that if there exist strong power system ground currents or electromagnetic fields then either metallic parts of the building and/or the underlying rock might produce sound and vibration by electro or magneto –seismic conversion, the latter which can also enter the building concerned.   Ground currents are typically caused by phase imbalance on power grids. Wind farms are notorious for causing such imbalance.


Thus the Hum might be expected to be more pronounced then either near to the source of pumped storage plants, near to wind turbines, near to industrial sites where there are increased levels of harmonics or at the end of a long transmission line under a large load.  Sometimes Hums may occur at other nodes in power systems. It is known that transmission line resonance factors can occur for fundamentals and harmonics even at line lengths as short as one eighth of one wavelength (Lundquist).    Improvements and new developments in both the communications and power fields have, for the most part, eliminated the need for transposing high-voltage power lines at close intervals. In many systems transpositions are made only at switchyards and substations. Geometric unbalance of un-transposed power lines leads to residual ground currents in solidly grounded systems. Changes in diurnal power flows will also change the Hum. Longer term changes in power flows across the entire country will also cause   sporadic outbreaks of the Hum in places which haven’t previously suffered and could also bring abrupt cessations of Hums in other places.



Taking specifically, for example, the Hum in Bangor and the surrounding area as an example, this maximises at night when Dinorwig is pumping. Dinorwig has been specifically documented as having oscillatory behaviour in its output under certain conditions and each of its 6 Francis reversible turbines behaves differently and is linked to a common water feed with some unwanted coupling taking place. Water hammer is also a feature.     Essentially under these conditions it represents an ‘industrial’ load on the end of a long transmission line.   Wind power also contributes to the problem.     The author has previously noted that the Hum, at least in Bangor, is worse when certain pairs of  Francis turbine pump units are operational at Dinorwig (refs)  and there is excellent evidence in the Scientific Literature as to why this should be with Munoz-Hermandez and Jones (2005) stating that ‘the behaviour of Diniorwig  is strongly dependent on the number of units active’ .      Dinorwig also has a physical vibration spectrum which includes the synchronous machine shaft speed of 8.3      Hz and its harmonics (Pritchard 1998).   



Another area specific feature of the electricity network along the North Wales Coast is the power flows at night and by day are radically different.  The super-grid and some Scottish Power transmission lines run substantially close and in parallel path.  There is the potential for unbalance due to induced negative phase sequence voltages.   



Perception of this kind of Hum requires no special bio-properties other than sensitive low frequency hearing and sensitivity to infrasound (Barnes).  Even though the prime mover is a common one different power grid frequency harmonics and sub –harmonics all subject to oscillatory and pulsation behaviour for the reasons above can combine by cochlear non-linearity to form the Hum.   By using three field spectrum analyses (Barnes) one can see the changes in real time which give rise to a particular Hum.  If two or more frequencies are beating together to give the Hum like effect or if the frequencies are buried in noise, so called stochastic resonance,

( conventional audio recording may not be possible.    Presumably this was the case with the original Taos Hum?  It should be stressed however that with increased computational power available to all since those days has come the ability to make virtual real time frequency spectrum analysis and present the results in time, amplitude, frequency colour waterfall display format which allows weekly pulsating audio signals in noise to be displayed. 






Perception of Hum by deaf people 


There are only anecdotal reports of this, nevertheless it is worthy of attempting to develop a theory to encompass such accounts.    The only feasible hypotheses would appear to be (electro) magnetic, gravitational, or depending on the type of deafness bone conducted sound/vibration.  


Such models of Hum perception may also enhance Hum in hearing people under conditions of stochastic resonance by so called synaesthesia or duelling of the senses.  In this case this would be duelling between conventional audition and a form of magnetic or gravitational audition.    At least for the present author who although not deaf has firsthand experience in Hum perception, the magnetic model would seem most feasible (Barnes) and has been tested in so far as an anomalous case of hearing Hum like effects in a vehicle not under power lines (Barnes) or at specific distances from pulsed radio transmitter antennas (Barnes).  It is not certain if all individuals possess this ability because it might be linked to ratios of biogenic and mineral magnetite in the body (Barnes).  



Perception of the Hum in buildings with Solar PV

In the above two cases the Hum would be expected to be perceived whether or not the local building electricity supply was energised.  The author has experienced the Hum in buildings with Solar PV.  Solar PV inverters can both receive from and transmit distorted waveforms to the mains. This   is so called embedded generation. The inverter itself may be the acoustic source and strong electromagnetic fields will surround it.  The Hum from these sources has the potential to be accentuated because the solar panels can also act as VHF/UHF radio antennas.   



Possibilities for mitigation/prevention


By better control technology it may be possible to modify power flow oscillations (  Before Japan had the Hum its main power flow oscillation frequency was of the order of tens of millihertz   and it is doubtful humans would perceive this.  One wonders if it could be an earthquake trigger however!


It is seen as highly unlikely that the amount of harmonics and pulses our power grids will diminish any time soon.  


The author is aware of two cases of the Hum, one in Charlbury Oxfordshire and one in Woodlands Tyne and Wear which appeared to start after ground-works.    It is postulated that these disturbed either channels for conductivity of power grid seismic signals in some way and/or ground conductivity and power grid earth current flows. 


Either way there is a potential route for R@D in Hum reduction/mitigation here.