Monday, 16 October 2017

Data Centres Vs Steel Plants - A Comparison


•  Steel plants use 60% of the energy demand of a data centre but provide 26 times as many jobs

• Port Talbot steel plant in Wales provides 28.5 jobs for every megawatt of demand compared to 1.5 at the data centre at Athenry 

• Approval for planning permission of any industrial project should require a high jobs to energy demand ratio of at least say six or seven. 

Last week, Apple received approval for their data centre in Athenry, County Galway to great fanfare in the media. I can only find one article (in the Independent) which dealt with facts (Revealed: Data centres to swallow 75pc of growth in Irish power demand). The article gives a good overview of the problems that lie ahead. 

I am not in favor of opposing something for the sake of opposing it. I'm in favor of discussing all the available facts and basing decisions on those. What we have now in Ireland is approving something for the sake of approving it which is just as bad as the other extreme.  I fail to understand how a country which prides itself on it's higher education credentials does not discuss the facts in relation to new projects such as data centres. 

David Hughes wrote previously on the Athenry data centre (The Cloud Bytes Back) :

To give an example Apple are seeking permission for a 240MW data centre in Athenry Co. Galway, which will create up to 215 jobs. The electricity consumption of this data centre will be the same as 420,000 Irish homes. This is ¼ of all Irish homes or every single house in Dublin City, Dun Laoghaire, Fingal and South County Dublin combined. Basically, the electricity needs of 1 Million people.
It seems the jobs figure has been revised downwards to 150 full time jobs according to the above Independent article.  So that's about 1.5 jobs for every megawatt of demand.  

Let's compare that to another high energy industrial user - steel plants in the UK. Port Talbot steel plant in Wales was due to shut down in the near future but the employees fought hard and the plant remains open for the time being. It employs nearly 4,000 people.

Port Talbot steelworks’ current demand for energy is about 140 Megawatts (MW), about half of which is internally generated. That works out at 28.5 jobs for every megawatt of demand.

A steel plant therefore generates about eighteen times more jobs per megawatt of demand than a data centre. Even if those jobs were cut in half by new technology, steel would still provide more jobs by a factor of nine.

Based on figures for Port Talbot then, steel plants use 60% of the energy demand of a data centre but provide 26 times as many jobs
The Government is keen to promote Ireland as being a location where it can meet the needs of the IT sector by providing certainty around planning and power supply
Interestingly, the government is not trying to promote Ireland as a location for steel plants which would create many times more jobs. 

The impact on energy demand, fossil fuel imports, emissions, electricity prices and 2020 targets will be enormous from data centres. As a consequence, it will be harder to attract other high energy industrial users that could provide many more jobs.  There needs to be a good payback for Ireland Inc. to compensate. Of course, they need to built somewhere to provide the demand for internet services.  But Ireland should not allow so many to be built here. We simply cannot afford it. 

Approval for planning permission of any industrial project should require a high jobs to energy demand ratio of at least say six or seven. 


Saturday, 14 October 2017

Electricity Retailers Increase Prices

Most electricity retailers are increasing their prices this month and the blame is been put on wholesale prices. However, gas prices are no higher than 2005 levels. 

Figure 1: Gas prices since 2000

I have been keeping track of my own electricity bills since 2012. I wanted to see if the reduction in wholesale prices and gas prices have been passed on to the consumer. The result can be seen in Figure 2.

Figure 2 shows little correlation between gas prices and unit price of electricity

The result is clear. The large fall in gas prices has not been matched by a similar fall in the unit price of electricity.   There has been a reduction in gas prices of about 50%. The reduction in the unit price of electricity has been about 7%. If we compare the energy payments or the annual market value of the electricity wholesale market with the gas prices we do see a better correlation (Figure 3).

Figure 3 shows good correlation between Annual Energy Payments (in orange) and Gas Prices(in red)

This means that when the gas prices are low, generators receive lower prices from the market. But these savings in wholesale prices are not passed on to the consumer in any meaningful way.

The hidden force in these graphs is wind energy. It has increased every year since 2012 now making up about 23% of the electricity mix. While we are constantly told that wind energy reduces the wholesale price of electricity, there is no evidence in the actual data.   With this increase in wind energy, there has been a parallel increase in system costs - grid and transmission infrastructure, back up costs, new interconnectors, and high wind penetration feasibility programmes (known as DS3). 

All these costs make up the unit price of electricity. After that, the PSO Levy gets added on, another component that only ever increases (an increase has also been announced this month).

The Energy Regulator seems to be toothless in the face of an electricity sector with out of control costs, government interference and regulation. His name is now redundant, and increasingly appears more like something out of Orwell's novel 1984 where Government Departments like the Ministry for Truth do the opposite of their name.       

Monday, 9 October 2017

Fossil Fuels Have Dramatically Increased Life Expectancy

From World in Data

Historic Usage of Fossil Fuels

The above graphs are self explanatory and do not require an explanation. As a species we have become obsessed with "carbon emissions" and "clean power" but we have at the same time completely missed the bigger picture - which is that since humans discovered fossil fuels as a power source they have lived for longer than ever before. 

It doesn't mean that pollution is not an issue, it just means it needs to be seen within the larger context of the incredible contribution of fossil fuels to humanity.

Club of Rome - I will just leave this link here for those who want to dig a bit deeper.

Sunday, 24 September 2017

Sea Level Rise (or Not) in Ireland ?

Climate change has caused the sea levels around Ireland to rise by almost 7cm since the early 1990s. This is due to the rising temperature of the planet, greenhouse gas emissions (GHGs) and the melting of glaciers. And the next five to 15 years is a crucial time period within which to act if we want to halt any permanent changes to the planet.  According to the Environmental Protection Agency (EPA), climate change is most obvious in our changing sea levels.  “Observed climate change impacts are most evident in the global temperature record, sea-level rise, loss of glaciers and ice sheets and changes in the nature and intensification of precipitation events,” the EPA report states.“  Since 1993, average sea level has risen around Ireland by just over 3cm per decade,” the report reads.  The agency said that Ireland is particularly “vulnerable” to climate change because we are an island country with major cities located close to coastal areas - Irish Examiner.

There are only two sets of sea level data available in Ireland that I can find - Malin Head and Dublin.

Malin Head shows a fall in sea level from 1958 to the early 2000s. This comes from PSMSL.ORG.

Malin Head

The part in red has been flagged with possible errors. But clearly, choosing to measure the trend from the early 1990s will show a small rise in sea level. But measuring from the beginning of the data i.e 1958 would show a fall in sea level.

So its interesting that the EPA have chosen 1993 to begin their sea level trend from.
 Since 1993, average sea level has risen around Ireland by just over 3cm per decade - EPA
After 2002, the Office of Public Works took over operation of Malin Head. Again, no upward trend can be seen in the data. In fact the data begins at 3.1m in 2004 and ends at just below 3.1m in 2011.

Malin Head 2004-2011

It's not clear how the sea level fell from approx 7m to 3m between the 2001 and 2004 graphs when the OPW took over operation. I presume that they changed the local reference height.

The other data available comes from Dublin. This data goes back to 1938 and goes up to 2009.

Here there was a very small rise in sea level :

The mean sea level trend is 0.07 millimeters/year with a 95% confidence
interval of +/- 0.42 mm/yr based on monthly mean sea level data from
1938 to 2009 which is equivalent to a change of 0.02 feet in 100 years.

So in 71 years there has been a sea level rise of 4.97mm - or about half a centimeter

This is a long way from the 7cm rise since the early 1990s as reported by the EPA and Irish Examiner.

This then leaves just the satellite data. This shows a rise of 7cm between 1993 and 2016 and a further 1cm in 2017. However, it's important to note that the satellites need to be calibrated using the tide gauges :

Sea level can also be retrieved with measurements from space-borne sensors. These sources indicate a sea-level rise around Ireland of approximately 2–3 mm per year since the early 1990s [till 2012]. However, satellite measurements need to be calibrated carefully, the most common method being the use of tide gauges. 


This raises the question - how can the satellite data and tide gauge data vary by so much ? Compare it with the most reliable tide gauge data - Malin Head from 2004 to 2011*. Malin Head is also perfect considering the satellite data is for the North Atlantic.

Comparison of Malin Head Sea Level with North Atlantic Satellite Data

Satellite 2004-2011

The satellite shows an overall increase of approx 1.5cm. 

Malin Head shows no discernible increase at all. Notice how Malin head shows variances of 0.1, 0.2 and 0.3 of a metre within the same year. That means the sea was rising and falling by as much as 30 centimeters in some years ! That is a rise of almost one foot ! 

By comparison, the largest variance within any given year in the satellite data is just 2 centimeters. 

I do not understand why the satellite does not pick up the seasonal sea level variations of one foot in a year which incidentally are confirmed by sea level data at Aberdeen in Scotland. This makes me suspicious of the satellite data. But if the satellite is correct, then how is Malin Head not showing any increase at all ? This would mean the reference point on the land would be shifting at the same rate (highly unlikely ?). 

All in all, the case is not closed on sea level rise in Ireland. It may well be the case that in the future scientists will make an announcement about problems with satellite sea data similar to the announcement they made last week about their temperature models.

Also, is 24 years a long enough period to draw any conclusion ?

*According to PSMSL for Dublin - After examining the historical files it was discovered that data for Dublin up to 2001 was supplied as Mean Tide Data and not Mean Sea Level data. Because of this the data has been flagged within the database.