Solving a case of the HUM by Dr Chris Barnes, Bangor Scientific and Educational Consultants,  Bangor Gwynedd, Wales, e-mail    Home:






The HUM is a general term for a peculiar and until recently unexplained or anomalous noise heard in recent times by sufferers, known as HUMMERS, the World over. The HUM is described by those afflicted as a very low frequency and pulsating or quasi-period sound rather like that of a distant idling engine. Musically minded HUMMERS seem to tone match the phenomenon in the range 30-80 Hz yet studies have often been done in locations prone to the HUM which fail to find any evidence of acoustic emissions at such frequencies (ref).     There have been several causative theories of the HUM ranging from electromagnetic audition through to physiological illness or abnormality. An excellent review of the HUM has been given by Deming (2004) who, however, then goes on to reach a rather peculiar and exotic conclusion that the HUM is most likely due to low frequency radio emissions from TACAMO      military aircraft.  The hypothesis advanced here is that the HUM is at least in part or in whole due to infrasound and that persons who experience the HUM have exceptionally sensitive low frequency hearing which extends down into the infrasonic range. Such sensitivity is not unknown (refs).



Bangor in Gwynedd, North Wales has its own particular case of the HUM.



The most credible theory of the HUM, namely that it is due to at least one infrasonic component working in concert with other acoustic frequencies has been advanced by Barnes (2007). Barnes (2007) has also shown that at least for some sensitive subjects living in the presence of electromagnetic transmissions possibly also enhances the sensitivity of the ear to low frequency sound and infrasound.   



Anomalous phenomena such as the HUM are notoriously difficult to do good science with since one has to rely mainly on anecdotal reports. However when their content contains numerous repeating and credible threads they are worthy of consideration for explanation.  Barnes (2007) and his experimental subjects have found the HUM intensity to be dependent on many variables as mentioned in some of these reports. Variables include; time of day, weather in particular wind speed and direction, high level winds or jet stream, nearby moving vehicles, location and lighting.  Further   infrasound   both in terms of its propagation and in human perception has been shown a most likely candidate to be dependent on such variables. Barnes (2007) has suggested that in North Wales a possible source of infrasound contributing to the HUM arises from the pumped storage electricity generation station at Dinorwig, near Llanberis, Gwynedd.     





Experiments to confirm Pumped Storage Scheme as source of infrasonic component of HUM.   


The experimental subjects are the author and his wife, both HUMMERS. The author has perceived the HUM for some two years now and his wife for four. The author has been logging instances of the HUM for some three years on a subjective scale of 0-10. Where 0= not perceptible and 10= seriously intolerable. Also logged have been times when the author’s sleep was disturbed as a result of the HUM.   The effects of wind speed and direction on the HUM have been determined. Recently archive data showing the operating times of the six reversible pump-generator sets at Dinorwig have been obtained (courtesy of  The data have been compared with times when the HUM was audible. Because the pumped storage scheme at Dinorwig is one of the mainstays in UK pumped storage generation there is rarely a day it is not in use. Such a day took place on 6th April 2007, when none of the pump-generators was active between midnight and 8am.  The archive log shows that neither the author nor his wife perceived the HUM until 8 am that morning.     



The bmreports website also shows predictive times when the pumped storage system is due to export electricity to the UK National Grid.  By using this feature it has become possible to listen for the HUM in advance. Given favourable still weather conditions, low instances of vehicular movement and no background household noise it has become possible to further test the hypothesis that the said power station does indeed produce the infrasonic component of the Bangor HUM. 



A first such experiment was conducted on 18th November 2007. Bmreports indicated that Dinorwig would commence exporting electricity to the UK grid at 1551 GMT. The author positioned himself in the quietest part of the house at 1545 GMT and waited listening intently but not looking at his watch. The HUM was not initially perceived. When the author perceived the HUM, reference to his watch showed the time to be 1551 GMT, strongly suggesting Dinorwig to be the source as predicted.  



A second double-blind experiment was performed on the night of 18th November 2007. The author obtained the the initial pumping time for the first generator set due to switch on in pump mode and retained it. The author’s wife particularly sensitive to the HUM was asked to listen out for it and report when it started. She was not told when to expect the HUM. Despite sitting in front of a TV set she became somewhat agitated at 2254 GMT and left the room to a quiet part of the house. On return at circa 2257 GMT she reported that she was hearing the HUM. Only at this sages was the data from bmreports disclosed to her, namely that the generating set known as T_DINO-4 had started pumping at precisely 2254 GMT. 


These experiments are in very strong support of the hypothesis that Dinorwig pumped storage scheme is responsible for one of the infrasonic HUM components in Bangor. 




Other retrospective re- examinations of the data.         


Various other re-examinations of the perceived HUM data have been made in comparison to bmreports.


The following are findings


1.      The HUM is worse when the station is pumping then when it is generating but this may due to diurnal sound propagation effects.    

2.      The HUM is worse when the station is pumping with 4 or 5 of its generating sets rather than all six. It is worst when 4 are in use.

3.      The HUM can also be heard across the Menai Straits in the village of Llanfair PG.

4.      The station usually seems to generate with one or two generating sets at any given time the HUM is worse with two in use rather than one.


Renewable energy sources in general are known to introduce more harmonics and disturbances into power grids.  These aspects of the Hum will be explored elsewhere.