China’s current power crisis will take just two years to be resolved, according to the country’s government. But the legacy of poor planning and conflicting policies mean that five years is a more realistic timescale.
As China struggles to resolve its latest power supply crisis, some 70 GW of new capacity must be built by the end of 2005 if it is to be resolved within the two-year time frame promised by the State Development and Reform Commission (SDRC).
But as this would require almost 30 GW more new capacity than the current fairly rigid Tenth Plan originally envisioned and as time is so short, a much more likely scenario within the context of a rapidly growing Chinese economy would entail up to five years of shortfalls in a significant number of provinces. Such shortages would manifest either directly as moderate to severe typically peak load rolling brownouts or blackouts, or indirectly as commensurate administrative controls on electricity consumption in specific sectors. Either way, it could be a long haul.
By the end of 2003, 22 of China’s 31 mainland provinces had experienced power shortages
This view is held to varying degrees by energy economists at the World Bank’s Beijing office, Tsinghua University, the Chinese Academy of Sciences and Japan’s Institute of Energy Economics (IEE). In all cases, they point to underlying structural difficulties that will prevent rapid crisis resolution. In essence, much more will be required than the quick fixes available to the SDRC but achieving the additional measures will take both time and greater macro-economic sophistication.
The crisis began in June 2003 when province after province began to suffer extended wide area peak period outages. By year-end, 22 of China’s 31 mainland provinces had experienced or were still experiencing extended rolling brownouts or blackouts with expectations that 2004 will be as bad if not worse. The chief cause has been widely identified as a severe shortfall in available generation capacity due to a cluster of interlocking conditions. These were:
- Unexpectedly rapid economic and hence power demand growth
- Coal supply shortages and hence higher coal prices
- Drought and high temperatures
The roots of the crisis stretch back to 1998 when a preceding extended power capacity new building programme throughout the 1990s clashed with a cyclical economic downturn to produce massive over-capacity.
The government’s response was to declare a three-year moratorium on new thermal power plant approvals that formally ended in 2001 while simultaneously vigourously promoting power consumption. Thus power tariffs to the electricity intensive aluminium smelting industry were reduced 17 per cent while urban households were encouraged to switch from coal to electricity and then install air conditioners, heaters and showers. Power networks were also extended to unserved rural areas while existing urban distribution grids were upgraded.
Such measures rapidly restored power demand growth. As Table 1 shows, final power consumption growth has accelerated every year since 1998 except for a slight dip in 2001. But as generation also increased to meet demand, the rate at which new capacity was installed decelerated each year to 2002 as the moratorium on approvals took effect. Such declines took capacity growth rates to well below GDP growth. Capacity installations only began to accelerate in 2003, far too late to meet the year’s anyway exceptional 15.3 per cent growth in demand. As the 2004 figures clearly show, this year looks significantly worse. Installed capacity growth will still lag drastically behind demand which in the first quarter to 31 March grew even faster than in 2003.
Two other factors exacerbated this already severe situation. First, as part of overall reform towards a market style economy, China has allowed coal prices to respond to market demand while still controlling kWh tariffs. This step initially caused no difficulties. But when most small coal mines that collectively used to supply up to 40 per cent of total coal were closed on environmental and safety grounds, its price began to rise. These increases then squeezed virtually all central and northern thermal power plants just when they should have been generating maximum output.
Second, severe drought and high summer temperatures simultaneously significantly reduced hydropower generation in several provinces while encouraging much greater than normal power consumption, particularly for farm irrigation.
Although the relative weight in the main crisis of these two factors is not yet clear, drought was a severe limiting factor in all the provinces whereas coal shortages were far more general. Rapid economic growth coupled with limited capacity availability, the main issue throughout the country, had a particularly severe impact in the provinces. Chongqing experienced both the fastest power demand growth in the country as well as hydropower shortfalls. Underlined provinces were not affected.
Table 2 provides a useful insight into the crisis. To begin with, it is by no means uniform across the country. In very broad geographical terms, the worst affected provinces form an arc that begins in eastern Jiangsu and then follows the coast south and west to curl up into the central provinces of Sichuan and Hubei. The crisis also relates to the very high maximum demand peaks experienced in all networks that were nevertheless smoothed by marked differences in trend growth in peak power demand.
Thus the north-eastern typically moderate power demand growth Dongbei and Huabei networks have so far been only moderately affected and Shandong not at all. Even so, the shortfall was sufficiently severe for factory owners in the Dongbei ‘rustbelt’ to welcome Russian offers of hydropower exports. There may also have been power imports from North Korea.
The more rapidly growing Sichuan, Nanfang, Xibei and Huazhong networks meanwhile experienced the full range of conditions with Huazhong apparently largely immune. It has access to more reserve capacity than any other network. Finally, very rapid power demand growth in the Fujian, Huadong and Chongqing networks (Chongqing actually belongs to the Sichuan network but has been differentiated to identify its remarkable load growth) means they are all severely affected. Given such growth, at least some shortfall might well have been unavoidable in any case.
Such a breakdown lends credence to SDRC claims that the crisis will be resolved in 2006. Not only could rain greatly mitigate power shortages in six provinces but the SDRC has also moved swiftly to rectify coal supply disruption. It thus increased effective 1 January the industrial retail power tariff by 8¢/kWh while also abolishing all preferential industrial tariffs. It has consequently at one stroke both improved power generation profitability and hence operations while also significantly curbing overall power demand growth. In 2002, the industrial sector absorbed 69.8 per cent of total final consumption.
More broadly, Chinese premier Wen Jiabao announced end April a series of forthcoming macro-economic controls on money and land supply and the pace of construction that should moderate the country’s boom-bust economic cycles. It is these that ultimately make power sector planning so difficult.
None of this addresses the immediate underlying problem of insufficient capacity in the context of continuing breakneck electrification. On the demand side, China must continue for powerful economic and environmental reasons both rapid economic growth and electrification whose combined impact on future load will be huge.
As Table 3 shows, China now resembles North Korea more closely than any of its other northeast Asian neighbours in both power consumption per unit of GDP and per capita consumption. It therefore has an enormous transition ahead of it. But since it must maintain high single digit annual GDP growth rates while simultaneously switching to electricity from other energy forms, particularly coal, annual average power demand growth must hover around ten per cent or higher over the next two decades. And since the transition to electricity typically takes off at around $1000 GDP/per capita, a level that China has just reached, near term actual annual load growth could well reach 12-13 per cent. In short, under the desired more orderly, more energy-efficient fast growth development pathway, power demand growth rates that would approach present high levels in some provinces would spread to all provinces. It is going to be an exciting ride.
In such circumstances, power supply cannot be allowed to fail. But the SDRC industrial tariff fix has only made high priced coal more palatable to power generators without addressing the underlying coal shortage. This will not be relieved by residential fuel switching to electricity because electrified households consume significantly more energy than they did from coal and 70 per cent of all China’s power is generated from coal. There can therefore be very little surplus coal production capacity that could fuel urgently needed new power stations.
The SDRC’s 3 January response was to announce more rapid construction of 13 major coal production bases around the country. But this will not only take time, it will also take the country straight back to precisely the same coal transport bottlenecks that have constantly constrained Chinese energy supply.
This unsettling possibility emerges from a comparison of three recent Chinese power demand forecasts (Table 4). Given actual demand of some 1900 TWh in 2003 and probably conservative forecasts of 2100 TWh and 2300 TWh, respectively, for 2004 and 2005, it can easily be seen that all three forecasts were far too low. The closest to reality was the ERI/SDRC (then SDPC) 2001 forecast showing 1900 TWh demand in 2005 and 2500 TWh in 2010.
But this forecast was scrutinized by Western Australia’s Murdoch University in September 2001 as part of an assessment of China’s likely future LNG demand. Murdoch pointed out that the forecast “is based on the assumption that China will maintain an elasticity of electricity demand [to GDP growth] of around 0.7 that is far below most developing countries. In a recent study conducted by the World Bank (2000, LNG in China), the elasticity of electricity demand for the period 1980-97 was 0.83 which was exceptionally low for any country. This shows that China will need to further increase its [forecast] electricity demand relative to GDP growth”.
China past, China future
The point could hardly be more timely. An increasingly electricity intensive but still consumption inefficient economy necessarily uses much more power per unit of GDP than either a lightly electrified one (China past) or an economy that is efficient (China future). China’s actual present coefficient may therefore be well over 1.0 per cent but until it plans accordingly its forecasts can only mislead.
Two possible power infrastructure development scenarios for the next few years emerge from all this. First, if the actual weight of drought and disrupted coal supply in the overall crisis was heavy, SDRC promises are credible and its planned new building will be sufficient. But second and unfortunately much more likely, if the underlying planning failure actually carries significantly more weight than the other factors, the crisis could continue in one form or another for several more years.
Not surprisingly, China’s energy industry has already reached the same conclusion. No less than seven LNG terminals are therefore now proposed along the coast for completion by 2010. The first three of these, at Guangdong, Fujian and Zhejiang are now certain with the fourth at Shanghai probable. The other three at Tianjin, Jiangsu and Shandong are all very much children of the crisis in the sense that their LNG power would have to be transported to relatively distant load centres.
But look at the timeline. New LNG-based power cannot be widely available for around five more years. Meanwhile Chinese plans for dramatically more nuclear capacity in much the same provinces could easily come on-stream in seven. So what is in prospect, is a fumbled shift away from coal in favour of LNG and nuclear – the corollary to which is around five years of misery in various provinces depending as much as anything on the pace of economic growth and the weather.