Electricity costs for smaller and larger households

Regular economics commentator Frances Woolley turns her sights to electricity in a recent posting (‘http://www.theglobeandmail.com/report-on-business/economy/economy-lab/frances-woolley/why-ontarians-arent-saving-money-with-smart-meters/article2190147/’, The Globe and Mail, 4 October 2011).  I enjoyed much of her argument, but I will ‘dig into’ one aspect – and by doing so, return to a theme I have picked up in the past in this blog.

More specifically, she writes:  ‘ … Ontario’s old regulated price plans favoured small households over larger families. Consumers paid a higher price for power if they used more than a certain amount each month. So, for example, a one-person household consuming 400 kWh per month paid 6.8 ¢/kWh under the May 2011 rate schedule, while a four-person household consuming 300 kWh per person paid, on average, 8 per cent more, or 7.35 ¢/kWh.’  I agree with the specific statements, but take issue with some of the details.

She is comparing a ‘summer’ consumption of 400 kWh (one-person household) in a month with one of 1,200 kWh in a month (four-person household).  And, yes, under the two-tier system, the difference in the ‘commodity cost’ would be as she writes:  for the smaller household, all 400 kWh would be under the ‘summer threshold’ of 600 kWh and thus 6.8 cents/kWh; while the first half (600 kWh) of the larger household’s electricity consumption would be at 6.8 cents/kWh, and the second half (600 kWh) at 7.9 cents/kWh, to give an average of 7.35 cents/kWh.  My issue, however, is that these are not the ‘costs’ to the consumer – or the price that would be paid.  Instead, one must look at the total charges.

Using Hydro Ottawa as the service territory in the Ontario Energy Board’s bill calculator (Prof. Woolley is at Carleton University), I put in the consumption patterns, and got numbers as follows:

Charge	One-person household	Four-person household
Consumption in a summer month	400 kWh	1,200 kWh
Commodity charges	$28.14	$91.46
Delivery	$22.81	$48.15
Regulatory charges	$3.09	$8.78
Debt retirement charge	$2.76	$8.28
Sub-total	$56.80	$156.67
HST	$7.38	$20.37
Sub-total	$64.18	$177.03
Clean Energy Benefit	($6.42)	($17.70)
TOTAL	$57.76	$159.33
Cost/kWh	14.44 cents	13.28 cents

Two points to be taken from this.  First, it is important to flag that ‘commodity charges’ – though important – are not the ‘cost’ to the consumer.  And second, differences in commodity charges will not necessarily translate (‘exactly’) into differences in these final costs to the consumer.  Indeed, in the example above, the difference in the ‘per kWh’ cost to the consumer is not ‘8%’ (the 7.35 over 6.8), but instead is really ‘MINUS 9%’ (8.7% results from 14.44 over 13.28) -- that is, the larger household actually has the smaller unit cost for electricity.  Given that 'fixed charges' (reflected in delivery charges) are relatively large, and many of them fixed, regardless of one's consumption, these can overwhelm 'per kWh' calculations.  Therefore, while Woolley’s attention to distributional impacts is most welcomed -- and the ways in which smaller households and larger households will be affected by a move to time-of-use pricing is worth investigating further -- we should reflect further upon whether the difference in commodity costs for smaller and larger households in the conventional 'two-tier' system really told us the full story. 

Kitchen table chats

The image of Tim Hudak speaking to a couple about their electricity bills in their kitchen – in a Toronto Sun article from 3 October 2011 entitled ‘Retired couple gives PC’s a boost’ – brought back memories of past photo-ops in Ontario electricity policy campaigning.  It would be interesting to compare and contrast such kitchen-table chatter.  The one that I wanted to pull out – but which I could not find ‘copyright-cleared’ on the web – was of Premier Ernie Eves doing something similar.  Though not during an election campaign, his visit to the home of Keshab Hardett and family in Mississauga – to announce his price freeze – is an image that is still in my mind almost ten years later.  When I went back and looked at Ontario newspapers on 12 November 2002 – and saw it on the front page of many – I can more fully understand why it is so ingrained.  In the next election, perhaps a pool could be formed to guess the first time such a moment might occur!  :>

Impacts upon seniors' and low-income households

To be honest, it is an article that I would not have expected to read in The National Post.  Entitled ‘Ontario’s Electric Circus’, and from 21 September 2011, interesting points about the relatively-modest differences in costs – to operate individual electric appliances in the home – are made.  These are similar to a blog posting of mine from last month.

The point to pick up on in this posting, however, follows from the article’s comment that:  ‘But there's little concrete evidence that time-of-use electricity rates disproportionately impact low-income families, seniors or otherwise at-risk groups - Mr. Hudak certainly hasn't offered any - and where such evidence can be found, it can be mitigated.’  I had the opportunity to supervise a Master’s student at the University of Waterloo a couple of years ago who investigated this issue.  In short (and the reader is directed to her thesis for full details), she found that seniors and low-income households (at least her sample in Milton, and at least as the on-, mid- and off-peak pricing existed before they were changed more recently) were a bit worse off under time-of-use rates (as compared to the traditional two-tier approach) in summer and a bit better off under time-of-use rates in the winter.  And, important to recognize, is that any differences were relatively modest.

PC elaboration of new residential pricing plans

There was quite a lot of coverage of Tim Hudak’s elaboration of how he would give residential customers ‘the choice’ with respect to time-of-use or flat rates.  The Toronto Sun, in an article entitled ‘Meters outsmart power users:  Hudak’ and dated 19 September 2011, reported it as follows:  ‘But Monday he elaborated on how prices would be different under a PC government: There would be a flat rate for the first 1,000 kilowatt hours and a higher rate for use after that.  “The flat rate is simply set as an average of the other (time of use) rates,” Hudak said.’  A check of the PC website could not provide confirmation and/or elaboration of this on a posting that explored the smart meter issue.  The Liberals, in their own subsequent press release, suggested that Hudak’s campaign team had issued a press release in clarification, but I could not find it.

But the Liberals were quick to jump on the issue, with a press release coming out saying that this would be a dramatic increase in costs to consumers.  The Liberal argument is that the PCs are talking about an ‘average’ of the three time-of-use rates – averaging 5.9, 8.9 and 10.7 gives you 8.5 cents per kWh.  Alternatively – and probably the case – if the average was in fact weighted to ‘time-of-use’, then the calculation would be a weighted average of 5.9 (64% weighted), 8.9 (18% weighted) and 10.5 (18% weighted), which is 7.268, or 7.3, cents per kWh.  This is higher than the present ‘lower-tier’ price of 6.8 cents per kWh, though lower than the ‘higher-tier’ price of 7.9 cents per kWh (which occurs across the summer threshold of 600 kWh). 

It will be interested to see how this is elaborated by the PCs.  ToU and two-tier are meant to be ‘impact-neutral’ to the average customer, and that is why you will find the 7.3 between the 6.8 and the 7.9.  As I have argued elsewhere in this blog, if you allow customers to choose whether they are on ToU or two-tier, then the rates for each would have to be higher than if everyone were on one or the other (because consumers will choose the ‘better one’ for them).   Another interesting elaboration will be with respect to the threshold – 1,000 kWh (mentioned by Hudak) is the winter threshold, but, as noted above, the summer one is lower.

This article also had interesting points about the impact of ToU upon peak consumption -- namely, a claim by Brad Duguid (Liberal MPP) that 'smart meters ... have already shaved 4% off demand during the peak consumption period' -- and I have addressed that in a previous posting.

August consumption and hot weather

The article entitled, ‘Hydro bills feeling the heat’ (Toronto Sun, 6 September 2011), inspired me to look at my own consumption through August.  Courtesy of my local utility’s webportal (Waterloo North Hydro), I was able to download what I concluded were ‘hourly-DST’ data.  Here, I give some results, and make some comparisons.

Below, in the table, are some core results.

time period	consumption (kWh)	percentage of total consumption	commodity cost ($)	percentage of total commodity cost
on-peak	210.03	18.5	22.47	27.1
mid-peak	199.47	17.6	17.75	21.4
off-peak	723.74	63.9	42.70	51.5
total	1,133.24	100	82.92	100

So, though we were often ‘at home’ during this period, our use ended up being very close to the standard 18/18/64 division that the Ontario Energy Board cites as ‘usual’ with regard to distribution of electricity across the three time periods.

Indeed, when I compare our ‘commodity costs’ to what they would have been if we were not on time-of-use rates (but were on the traditional, ‘two-tier system’), I find it striking – and I report it in the table below – that the difference is one cent.

consumption bracket	consumption (kWh)	commodity cost ($)
below threshold (6.8 cents per kWh)	600	40.80
above threshold (7.9 cents per kWh)	533.24	42.13
total	1,133.24	82.93

Our consumption was high in August – indeed, a little bit higher than usual.  (For some historical data, see an earlier blog posting.)

In any case, I was interested to see how consumption correlated with temperature.  To explore this, I retrieved temperature data from the University of Waterloo weather station, and compared them with my hourly consumption data in three ways:  plotting daily average temperature versus daily consumption; plotting daily maximum temperature versus daily consumption; and plotting hourly temperature versus hourly consumption.  Results are below.

Average daily temperature versus daily electricity consumption,

single home in Waterloo, ON,

August 2011

Maximum daily temp. versus daily electricity consumption,

single home in Waterloo, ON,

August 2011

Hourly temperature versus hourly electricity consumption,

single home in Waterloo, ON,

August 2011

Though the visual suggests that the relationship is not completely ‘direct’ (that is, higher temperature, higher consumption), but the trend is there.  Indeed, in these three graphs, there are positive correlations, with values of 0.277, 0.120 and 0.378, respectively.  Hot weather, not surprisingly, leads to greater consumption.

7-9pm: decreased demand or political posturing?

I was curious about the extent to which the relatively recent change in the 7-9pm period (in the time-of-use environment) from ‘mid-peak’ to ‘off-peak’ was justified.  In other words, to what extent did demand drop during this time, or, alternatively, to what extent was it a departure from efforts to represent some kind of ‘real-time market’ in electricity in the name of ‘good politics’?

To investigate further, I pulled out some recent data – namely, for the eight weeks from 2 July to 26 August 2011, I collected, from the IESO website, the non-holiday weekdays (all 39 of them), and examined average ‘spot market price’ (Hourly Ontario Energy Price) and average Ontario demand.  These results are presented in the two graphs below, with the two hours of interest (7-9pm, which are ‘hours ending 8pm and 9pm in IESO-speak) circled in each.

Visual observation of these graphs suggest that while there is some ‘drop-off’ from the late afternoon peaks, the values still seem ‘moderate’, suggesting that ‘mid-peak’ may well have been right.  To look at it another way, I took the average values for each of these 24 hours, colour-coded them (red for on-peak, yellow for mid-peak and green for off-peak) and then ranked them from ‘top’ to ‘bottom’.  That graph appears below.  That graph further suggests that the 7-9pm period – at least on the basis of this limited investigation – might more rightly be called a ‘mid-peak period’.

Consumer choice of time-of-use or flat-rate

A 23 August 2011 article, entitled ‘PC candidate says smart meters bad for business’, appeared in The Barrie Advance.  In it, the PC Party’s plan for moving away from obligatory time-of-use rates was noted.  The relevant part of the article is as follows:  ‘[PC candidate Rod] Jackson said families and business should be able to choose what works for them – and opt out of smart meters. He couldn’t say, however, what rate they’d pay – an off-peak, mid-peak or on-peak rate. “It’s a good question. It’s up to the Ontario Energy Board. It won’t be as much as for a smart meter,” he said.’

I am left to wonder how prices can really go down when consumers are left with a choice of time-of-use rates or, say, some kind of flat-rate.  The revenues from the customers need, of course, to cover the costs of providing the electricity, so without a change in the cost of supply (which, admittedly, may be part of the broader PC platform), the money coming in has to be the same.

To give this further substance, I laid out the following simplistic example. (And noting my earlier comments in this blog, I recognize my own problem for concentrating upon ‘commodity costs’, but the point does not depend on that.)

Imagine customer A and customer B with load profiles as below.  If both are paying in a time-of-use environment, then the total revenue is 141.4 cents.  If both were paying in a flat-rate environment, the total revenue would be 140.0 cents.

Now if each was given the choice, customer A would stay on time-of-use rates (a heavy off-peak user) and customer B would choose the flat-rate tariff. Total revenue is now 136.8 cents.  Would it not be that the missing revenue would have to be made up by increasing the relative costs (say to 6.1/9.1/10.9 and 7.2 – that is, increases of about 3%), which would then generate revenue of 140.8 cents?

Customer	kWh use during off-peak (5.9 cents/kWh)	kWh use during mid-peak (8.9 cents/kWh)	kWh use during on-peak (10.7 cents/kWh)	Total cost under time of use (cents)	Total cost under flat-rate (assuming 7.0 cents/kWh à blending 6.8 and 7.9 cents/kWh prices)
A	8	1	1	66.8	70
B	6	2	2	74.6	70

It will be interesting to see if these options that will be available to customers will be more fully elaborated during the campaign.

Residential price increases since 2003

A longer story involving the early-morning electricity-user (noted in my posting below) also appeared on 15 August 2011 – this time on InsideToronto.com in an article entitled ‘Tim Hudak vows to slash hydro rates if elected’.  Two passages are worth noting.

First, there is a reference to one resident’s electricity consumption:  ‘"I get up in the middle of the night to do my cooking and to wash my clothes. I never use the dryer, I hang things outside to dry. I try and economize however I can, but it doesn't make a difference," she said Monday, pulling out her March electricity and heating bill totalling $1,698.85 - including $211 in Harmonized Sales Tax (HST) and $175 worth of Debt Retirement Charges. "I believe that's unreasonably high. It's a lot of money and a change would certainly be welcome."’   What was striking in this is that $175 in Debt Retirement Charges (DRC) suggests incredibly-large electricity usage!  The DRC is $7 per 1000 kWh.  Therefore, this cost suggests a consumption – in the month of March (?) – of 25,000 kWh.  Yes, it appears that she has electric heating as well, but it should be noted that, as a monthly consumption, this is large.  (The provincial average is about 800 kWh.)  Indeed, in work that we did recently, we looked at 1,020 households  and we found that, from this random sample, the largest user consumed 31,412 kWh IN ALL OF 2008 (average of 2,618 kWh a month).  This resident’s experience is unusual.  (Another interesting note is that we found – tentatively – that large users are ‘better off’ under time-of-use rates, as compared with the conventional two-tier rates.)

And, second, it was noted that:  ‘According to Hudak, hydro rates in Ontario have increased eight times by a total of 84 per cent since 2003, …’.  In a recent posting, I noted that my experience was an increase of about 30 per cent.  I wondered where the figure of 84 per cent came from?  I went to the historical prices in Ontario – solely for commodity (electricity) charges – and found that if you compare the higher tier price after 1 May 2011 (7.9 cents per kWh) with the government rate after 9 December 2002 (which was then the price throughout 2003) (4.3 cents per kWh), then you get a figure of 83.7%.  I think that that must be the source.  It is, therefore, a misleading comparison for at least two reasons:

1)      This is only the commodity charge, and the other (delivery, etc.) charges should be included to reflect the ‘final price’ to consumers.

2)      The ‘new cost’ (2011) only applies to those units of electricity consumed after the lowe-user threshold has been reached (600 kWh in the summer and 1,000 kWh in the winter).  Perhaps a blended value would be more correct.

And, of course, prices, as I noted earlier, were much higher in 2002 before parts of the market were closed.

When are those time-of-use rates, anyway?

I wonder how much confusion there is regarding ‘time-of-use rates’.  I ask, because of a passage I saw in an article entitled ‘Hudak steers clear of hydro price promises’  (Toronto Star, 15 August 2011).  The passage is:  ‘At the campaign-style stop, Hudak was met by local senior Li Zern who was complaining about her hydro bills.  Despite the fact that peak hydro rates end in the early evening, Zern said she is getting up “between two and three” in the morning to cook and clean to try to keep her electricity bills under control. “They’re going up considering that I’m economizing. I’m getting up in the middle of the night to do my work, even my vacuuming,” Zern, a Conservative party member, told reporters.’

Off-peak rates exist between 7am and 7pm, year-round, and all day on weekends and holidays.  I wonder the extent to which these kinds of reports – the ‘need’ to get up in the middle of the night to do ‘electrical jobs’ – become ‘urban myths’.  If they do, efforts to have an informed discussion about electricity policy may be hindered. 

Will Peak Pricing Create New Peak Periods?

On Wednesday, 10 August, an article by local radio commentator Brian Bourke, entitled ‘Confessions of a Hydro Tyrant’, appeared in The Waterloo Chronicle.  What jumped out at me was a theme that seems to be appearing repeatedly in the press – namely, that large shifts by users (in a time-of-use environment) will cause the on-/mid-/off-peak rates to be meaningless (or at least no longer matched with system-wide demand).   The quotation from the article is as follows: ‘Although, my theory says if everyone moves to the off-peak time, doesn’t that make it the peak time?’

A number of questions are raised by this.  One is – and I will approach it in a subsequent entry – ‘to what extent do the pricing periods match broader demand (and/or broader market prices)?  In other words, are we being charged ‘peak prices’ at the ‘toughest times’ and so on?  Again, I will look at that one in the near future.

For now, let me consider, more closely, an issue I have already touched upon in this blog – namely, the extent to which customers will shift their loads because of the changes in prices.  My ‘gut feeling’ is that the shifts in loads will be relatively small – or at least not big enough to have on- and off-peak periods exchange places – for two reasons.  First, the differences in costs are relatively modest.  As my entry entitled ‘On-peak and off-peak residential pricing’ lays out, the difference in cost from having the oven on for one hour at 6pm in the winter versus 8pm in the winter is 15 cents.  For the most part, other household dynamics will be key determinants in deciding when dinner is.  And second, not all electrical services in the household are time-flexible – you cannot, for instance, turn on and off your refrigerator, nor can you run your dishwasher before you have your meal.  Indeed, back in August 2004, I had an article entitled ‘Real-Time Pricing for Electricity:  What Difference Could it Make to Ontario Households?’ in Municipal World.  There, I published a hypothetical household’s monthly electricity consumption (1,000 kWh … the figure would be lower today), and highlighted the fact that different electrical services had different levels of time-flexibility.  That figure looked like this:

Used all the time	Used with some flexibility (regarding time of use)	Used virtually any time
Refrigerator (100 kWh) Chest freezer (100 kWh) Clocks (10 kWh)	Furnace fan or central air conditioning (300 kWh) Oven (100 kWh) Lighting (90 kWh) Dehumidifier or heat exchanger or heat-recovery ventilator (100 kWh) TV, microwave, coffee maker, computers, etc. (85 kWh)	Washing machine and dryer (90 kWh) Dishwasher (15 kWh)

Some of the values can be challenged, but it remains the case that not all electrical services can be used at whatever time.

More academically, a number have examined the so-called ‘price elasticity of demand’ with respect to electricity – that is, for every percentage increase in costs, what is the percentage decrease in use?  In the short-term, it is often argued that – in households – that value is something of the order of -0.2 or -0.3.  In other words, for every 10% increase in prices, consumption falls by 2% or 3%.  In the longer-term, as householders make investment choices to reflect new pricing realities (for instance), that value may be up towards -0.7, -0.8 or -0.9.  Recent investigations may be found a recent article in The Electricity Journal (‘Residential and Regional Electricity Consumption in the U.S. and EU: How Much Will Higher Prices Reduce CO2 Emissions?’, by Inês M. Lima Azevedo, M. Granger Morgan and Lester Lave, 24(1), January 2011) and in a recent CD Howe Institute Backgrounder (here (pp. 4-5)).  There is also a debate about the extent to which customers respond to (and understand) marginal costs (what they are paying for that ‘last unit’ – for instance, in a time-of-use environment) or average costs.

Some of the most comprehensive work on this issue has been done by Ahmad Faruqui in the United States.  His findings – see, for example, a recent summary article he co-authored in Public Utilities Fortnightly (January 2010), entitled ‘Rethinking Prices’ – suggest that our situation, with its relatively low on-/off-peak pricing ratio in a time-of-use environment, might yield a 10% reduction in on-peak consumption.   With the average household consuming 18% of its monthly 800 kWh during peak periods (or 144 kWh), then reductions will be of the order of 14.4 kWh a month – equivalent (to keep with an earlier example) to moving dinner once a week from 6pm to 8pm (in the wintertime) or not running the air-conditioner between 12 noon and 4pm once a month (in the summertime … that calculation from the IESO’s 10 Smart Meter Lane).  Research – some which we here at the University of Waterloo hope to undertake – will reveal what people actually end up doing.