Impact assessment of climate policy on Poland's power sector

Formula (1) differs from the methodology provided in the IPCC Guidelines (IPCC 2006). It was decided to include not only the electrical energy production but also heat, because it well suits CHP. The default emission data are given in literature, but the aim here was to obtain actual values for different technologies applied in Poland.

Balances of revenues and costs were done for all the groups. The ratio of the difference between revenues and costs to the energy production was defined in order to enable the comparison of the performance of each technology and fuel.

Profitability was determined by dividing total profit/loss on sales of electrical energy and heat with financial and other revenues and costs considered by the sum of revenues generated by the unit.

where P/L₃__{CO2 excl} stands for profit/loss on sales of electrical energy and heat with financial and other revenues and costs considered, excluding the impact of the EU ETS.

Profitability without CO₂ trading was counted in order to show the impact of the EU ETS on the power business.

There is a strong correlation between the energy produced and emission in each of the groups—the absolute CO₂ emission depends on the fuel used and the energy produced (Fig. 4).

The development of high-efficient CHP is one of the pillars of the EU energy efficiency policy (Directive EED 2012). What matters is that using hard coal in CHP plants can cut the emission factor by half in comparison with a power plant producing only electrical energy. It should be considered while exploring new investment opportunities especially in times when the EUA prices are inevitably to rise.

The emission factors reveal substantial, however expected, differences among power technologies used in Poland (Fig. 5). The most emissive fuel is lignite—over 1 t CO₂/MWh. In contrast, the least emissive are gas-fuelled CHP plants. In Poland, the number of gas-fuelled CHP plants is small due to a high gas/coal price ratio, and they are mostly used as peak capacity. Their environmental benefits as compared to other technologies have not yet been properly valued, partly due to low EAU-related costs. The situation is likely to change since there are several gas-fuelled plants under consideration following the need to use possible oversupply of gas under import contracts. Poland’s national inventory report covering all the EU ETS energy installations gives the average emissivity factor of EU ETS power installations of 0.8 t CO₂/MWh (KOBIZE 2015). The figure is very close to the results obtained in our analysis; however, our results distinguish between different technologies.

A large improvement in emission intensity from public electricity and heat generation in the EU-28 was observed in 1990–2012; as in 2012, it was 0.566 t CO₂/MWh compared to 0.754 t CO₂/MWh in 1990. In 2000–2010, the CO₂ emissivity fell by 29% (on average 1.7% per year). Since 2010, however, CO₂ emission intensity has increased by 2.6% per year due to a larger share of coal and lignite temporarily replacing gas. Comparing EU-28 and Poland’s average emission intensities in 2013, one can find 0.558 t CO₂/MWh and 0.713 t CO₂/MWh, respectively, which is unexpectedly a good result (1:1.28).

On the list of top 30 emitters in 2015 in the category combustion plants, there are four Polish power plants: Bełchatów with annual emission 37.1 Mt CO₂ and no change against 2014 (0%), Kozienice 11.4 Mt CO₂ (+ 4%), Turów 7.6 Mt CO₂ (− 7%) and Rybnik 6.5 Mt CO₂ (− 10%) (EEA 2016a). These not improving emission records trigger public and political criticism since the power plants are recipients of free EUA allocations under Article 10c. Especially, the case of the Polish Bełchatów power plant is widely publicised (CEE 2016; EC 2015b)^. It is claimed that the funds allocated to the power plant are used for restoring coal capacity instead of reducing GHG emission. There are also threatening voices calling lawmakers for excluding coal power plants from the enabled by the EU ETS reform energy transition funds. In contrast, in the World Bank’s RISE ranking (World Bank 2017c) embracing regulatory indicators for sustainable energy overall score, Poland with 78 points is globally on the 24th position and 14th in the EU (energy access—100, renewable energy—78, energy efficiency—57). However, the ranking overvalues countries with full access to energy and, therefore, is not fully suitable to asses highly developed countries.

In PRIMES model, carbon intensity of power generation from thermal plants steadily decreases: 0.41 t CO₂/MWh in 2020, 0.33 t CO₂/MWh in 2030 and 0.16 t CO₂/MWh in 2050. It is accompanied by a corresponding rise in efficiency of thermal electricity production—38.4% in 2010, 40.8% in 2020, 42.7% in 2030 and 44.6% in 2050. This diminishing trend is not reflected in Poland’s power sector. The average efficiency of Polish thermal power plant was 33.7% in 2011 and 34.2% in 2013 that was below the EU average—38.5 and 36.3% respectively in these years (WEC 2016). Only the new coal units planned for commissioning till 2020 will exceed the current EU average, e.g. in Opole 45.5% net (2 × 900 MW), in Kozienice 45.59% (1075 MW), in Jaworzno 45.9% (910 MW) and in Turów 42% (450 MW). It can be assumed that Poland’s energy sector has now reached stage I of the Clean Coal development, i.e. being retrofitted and with new capacity being in line with current standards of energy efficiency and emission reduction of SO₂, NO_x and dust requirements. Further progress is limited since achieving energy efficiency greater than 50% in coal-fired technologies still requires research especially when it is to be combined with carbon dioxide capture and storage (CCS) i.e. Clean Coal Stage III (Kavouridisa and Koukouzas 2008). Bryant (2016) discussed the impact of high coal dependence in highly concentrated power sector in the EU in 2005–2012 on socio-spatial concentration and centralisation of capital and GHG emission (EC 2015b). In the case of Poland, he concludes that ownership relationship changes supported concentration of GHG emission under the governmental control. The process has recently been even deepened as some of the coal-fired power stations were sold back (indirectly nationalised) by international energy companies like Vattenfall or Électricité de France (EDF) to state-controlled power companies. This makes the government particularly strong and potentially an effective enabler in the GHG emission abatement policy. It also puts extraordinary responsibility on the government.

Participation in the EU ETS compels installation operators to preserve their emissions at some predefined level. The comparison of the allocated EUA and the emissions verified by the EC over all the EU ETS phases is of interest to see whether the plants were able to make profit by selling over-allocated EUA or they had to buy the missing ones. It was done using data from the EU Transaction Log (EUTL). This database provides all the information on the individual power plants; therefore, it was possible to assign verified individual emissions into the established groups. The results show the extent to which the situation of the plants changed with the commencement of the third trading phase (2013) and the alteration of rules of allocation, e.g. from free allocation to auctioning (Fig. 6).

While in the first trading phase (2005–2007) the majority of the groups had experienced over-allocation which probably resulted in additional revenues, in the subsequent second trading period (2008–2012), the total EUA allocated remained sufficient only in the case of CHP plants fuelled with hard coal and until 2011 in units fuelled with gas. At the same period, EUA shortages experienced by other groups were not substantial. The beginning of the third trading period (2013–2020) changed the situation completely. In all cases, the shortage is getting higher every year, and the next years are rather not likely to bring any relief providing that the freely allocated share is to completely diminish by 2020.

Comparing Poland’s rate of emission from combustion over 2005–2014 period with other three main EU polluters (Germany, the UK, Italy), one can conclude that Poland has stabilised its emission while other countries show a certain degree of decrease (EEA 2016a). It can suggest that despite a growing share of RES in energy mix and efforts to reduce emission from individual power units by improving energy efficiency, a kind of stagnation caused by tapping all low-cost energy saving potential can be observed in Poland’s power sector.

Costs of production of electrical energy and/or heat differ with the technology and fuel used. In 2008–2014, the total costs ranged between €60/MWh and €27/MWh. The highest values were for hard coal power plants and gas CHP plants in 2012 (as gas is the most expensive fuel in Poland), and the lowest for hard coal CHP plants in 2009 (Fig. 7). The average production cost of electrical energy in coal-fired power plants was €44/MWh in 2014 and €46/MWh in 2015, i.e. 2.8% higher (URE 2017). Just to compare, the wholesale electrical energy sold on the competitive market of the Polish Power Exchange was lower—€42/MWh in 2014 and €41/MWh in 2015.

For the purpose of this study, the total cost and revenues of energy production consist of the following:

The costs of participation in the EU ETS categorised in the accountancy of power plants as a part of “other costs” were filtered from the financial data available. They vary between 0.2 and 11.5% of the total costs of production in the assessed groups. The biggest fraction of the CO₂ costs when compared to other costs is in the case of lignite-powered power plants (up to 11.5% in 2013) and hard coal power plants (up to 5.7% in 2013). They are quite acceptable and stable for CHP plants since they barely exceed 2% (Fig. 8).

The costs and revenues due to participation in the EU ETS were counted and expressed for one energy unit (MWh) of electrical energy and heat generated (Fig. 9). As expected, additional costs and revenues are considerably smaller for CHP technology, while the costs borne by the power plants are substantially higher. In the second trading period, some units were able to sell remaining EUA thus generated additional revenue—they benefited from the EU ETS participation, whereas it was never the case for lignite power plants and rare for power plants fuelled with hard coal.

The situation radically changed with the start of the 3rd trading phase as from 2013 only a fracture of EUA remained allocated freely. All shortages had to be covered by EUA purchase at auctions which increased the operational costs. Notably, in 2013 the costs generated by EUA purchase were higher than in 2014 due to a growing oversupply of EUA and the following drop of prices. It might also be a sign that power sector, noticing that the price of allowances started rising in May 2013, decided to take preparatory steps for the decrease of the number of freely allocated allowances and purchased too many EUA. In the extreme case of the hard coal power plants it resulted in €4 /MWh increase of production costs. However, the situation changed after the 3rd trading phase had commenced as in 2013 the period of growing losses began and it is likely to last until 2020 and afterwards. The losses are likely to steadily increase unless the power sector undertakes provisions to reduce its GHG emission.

Yearly profitability, measured as a ratio of financial profit (profit/loss count) and total revenues, was calculated for each fuel/technology group (Fig. 10). In the period of 2009–2012 for hard coal CHP plants and in 2012 for hard coal power plants, the profitability with the EU ETS excluded was lower, than that in the non-ETS case. That means that the participation in the system allowed the units (seen as a group) to gain an additional profit. The reason for this situation was that the number of freely allocated EUA that they obtained was bigger than their actual need. In all remaining cases, the profitability was either slightly (CHP-gas) or significantly higher (PP-lignite) with the exclusion of the EU ETS impact.

The power capacity and energy production till 2020 were estimated according to governmental projections (Ministry of Economy 2009; Ministry of Energy 2015). The power trend line was built to predict emission levels in the period of 2015–2020. To estimate the future emissions for the predefined groups, trends based on the data from 2008 to 2014 were used (Fig. 11).

The total emission of the groups will decrease by 13.4 million tonnes of CO₂ when comparing levels in 2020 and 2008. The decrease in the third trading phase alone will be of 5.5 million tonnes (Fig. 12).

The free EUA are individually assigned to power units according to governmental regulation (Fig. 13) although each year, the amount is adjusted according to the progress made in compliance with the planned investment. Over the third trading period, Poland is likely to obtain 424 million of free EUA for the whole power sector (Cabinet of Ministers 2014) as a part of the derogation mechanisms.

The future amount of EUA obtained through derogations of Article 10c had to be predicted. Historical data available covered only 2013 and 2014, the period too short to estimate the trend. Moreover, the use of free EUA in 2013 and 2014 was below schedule in all the predefined groups with growing tendency to diverge. Thus, it was assumed that the compliance for the whole period of 2015–2020 in each group separately would be constant and equal the average from the years 2013–2014, i.e. CHP-gas—84%, CHP-coal—86%, PP-coal—80% and PP-lignite—84%. However, later it turned out that the estimation was too optimistic since the actual share of all free allocated EUA used under Article 10c by power sector in 2013–2015 amounted only to 40% which was the lowest score in all eligible MSs (EEA 2016a). Despite the fact, the assumption still seems reasonable since the refurbishment and modernisation of power units require substantial investments which will force companies to struggle to maximise the share of free EUA and to accelerate investments at the end of the eligible period.

After forecasting emissions and freely allocated EUA in the years 2015–2020, the shortage of EUA for the groups was estimated. Considering the previously set EUA price scenarios forecasts, the estimation of the costs generated by participation in the system was made. The predicted shortage of allowances was simply determined by subtracting the free allowances to be obtained in the period of 2015–2020 from the forecasted emission levels (Table 2).

Projected costs of purchasing the missing EUA were counted by multiplying the yearly estimated shortage of EUA for a certain group by the EUA cost for the three scenarios (see Fig. 2). In order to determine a possible economic impact of the EU ETS participation, the projected costs were compared with the average historical, i.e. from the years 2008–2014, financial balances of the power plants. Since in this analysis the BAU scenario of the power plants was assumed, hence possible changes of other factors, e.g. fuel price, GDP fluctuations, energy demand, annual temperature change and technological progress, were not taken into account. It is possible that the EU ETS costs themselves reach the balance and thus make the conventional power plants not profitable (Fig. 14). The gas CHP plants are the least impacted by rising CO₂ costs. For the other groups, the base scenario costs will exceed the average balance in 2019.

When it comes to hard coal-fuelled plants, the difference between plants producing only electrical energy and CHP plants is noteworthy. The plants producing electric energy only proved a lot of inefficiencies. Their profitability levels were the lowest when compared with other groups and even reached the level below zero in 2013. When analysing the future, in regard to growing EU ETS costs, they showed the biggest vulnerability, and regardless of the EUA price scenarios, they may lose their profitability the soonest unless the energy costs increase. An assessment, similar to our short-term forecast, on profitability of coal-fired fleet in Texas (USA) concludes that fundamental changes in the Texas electricity market are putting coal-fired power plants under increasing economic and financial stress (IEEFA 2016b).

The second category, i.e. CHP, managed to gain some additional revenue in the past years as it is not a big emitter and, assuming minor changes in their BAU scenario, is going to remain profitable in the next years. The results support the conclusion that gas-fuelled CHP plants, now the smallest installed capacity in conventional power sector, are the sources best immune to price changing regarding the EU ETS-related costs. Even the most severe scenario (MAX) of EUA price change will not affect its profitability to the extent that the CHP plants lose economic viability. They are also very operationally flexible, which is of great importance considering the growing share of intermittent RES and the lack of significant energy storage capacities in the power system, e.g. pump-storage plants. The only drawback is that they use imported fuel and, therefore, they make the country more dependent on import. However, it is certain that its share in the Polish energy mix should be increased especially given that the gas does not need to be entirely of Russian origin anymore, since the new LNG gas terminal in Świnoujście was put in operation in 2015. It takes LNG from the Middle East, mainly Qatar, but in June 2017 for the first time from the USA what shows a growing openness and flexibility in global energy markets.

Moreover, lignite-fuelled power plants should not be phased out, because even though they demonstrate the highest emissivity, they are very cost-effective. Their average profitability over the years 2008–2014 was the highest when compared with that in the other groups discussed. The analysis of the impact of rising EUA price on the profitability of these sources showed that assuming a base scenario of the price change, they would lose their economic sense in 2019; however, with the minimum (MIN) scenario, they would remain profitable. This conclusion coincides closely with the findings of Climate Analytics (2017) which demonstrates that market-forced shut-down time of lignite power plants in the EU is longer than for hard coal plants due to the higher profitability of the lignite ones. For example, the report predicts for Bełchatów (5 GW on lignite) the shut-down time for 2055 and market shut-down for 2027. Additionally, the lignite mines are owned by power generators what makes their economic situation better as compared to hard coal sector still mainly controlled by the state. However, the lignite sector faces a problem of fuel scarcity after 2030 when the existing lignite fields will be partly close to exhaustion that may lead to a drastic capacity reduction from current 9000 MW to around a third (Climate Analytics 2017).

Wholesale electricity prices in the EU reached maximum in the third quarter of 2008 and, apart from a minor increase in 2011, have been on the slope ever since. Prices have diminished by almost 70% since 2008 and by 55% since 2015 and in 2016 recorded the lowest level for 12 years. This diminishing trend is not however reflected in the retail prices which exhibit a rising tendency. It is explained by steadily diminishing “energy” component in the total prices which fell by 2.8% in the period of 2008–2015 whereas the” network” component rose annually by 3.2% in these years, and there was a significant increase of the share of taxes and levies from 12 to 32% of the total price. In Poland, “energy” costs were slightly higher than “network” costs (EC 2016e). The tendency is growing since RES component and costs related to stranded assets in the power sector are steadily increasing not being fully compensated by the diminishing costs of energy generation. The two opposite trends make the energy prices stable over the period of 2012–2017. The wholesale electricity prices in Poland count to the lowest in the EU. Comparing electricity prices for end users in Poland, it can be noticed that they are below the EU average and much lower as compared to the most expensive MSs. The price for industry in the second half of 2015 was €86/MWh whereas the EU average was €119/MWh and the peak in Italy €160/MWh and the UK €149/MWh. The price for households is controlled by the energy regulator and kept low to protect citizens, and it was €198/MWh, which is lower than the EU average €211/MWh and substantial lower than in Denmark €304/MWh and Germany €295/MWh. Therefore, the Polish society apparently benefits from the extensive use of coal in power generation. It cannot be forgotten that not all public subsidies and tax reduction in the energy sector are publically available which makes the economic assessment of actual energy costs highly uncertain. For example, data from the Organization for Economic Cooperation and Development show that Poland’s government supported its coal industry with more than $840 million in tax expenditures and budgetary support in 2011. Consumer and producer subsidies would make the number higher (Kowalski 2016). On the other hand, it is notable that according to the official EC database (EC 2016d), the state-aid expenditures (tax and fiscal measures) in Poland were one of the lowest in the EU amounting to €1 billion cumulative in 2008–2014, hardly comparable with Germany public support, in the form of tax allowances, tax advantages, direct grants and tax base or rate reductions, surging to €42 billion. This disproportion places Poland’s energy sector in much unfavourable, less competitive position against our neighbour. The EC warns that in the North-Western Europe region (Central and Western Europe, Nordpool and the UK), natural gas prices and emission allowance prices (ETS) had stronger impact on the wholesale price level than in Europe as a whole (EC 2016d; OECD/IEA 2014). The EU permanently experienced a high price ratio of gas to imported coal as it was 1.7 on average in 2008–2009, amounting to 2.7 in 2014–2015 and then falling to above 2 in the beginning of 2016. This decreases the share of gas in the European energy mix, although being accompanied by a diminishing share of coal gradually replaced by cheaper RES technologies (EC 2016d; OECD/IEA 2014).

Let us analyse some factors affecting energy prices in Poland. The first is the coal price. Due to the high proportion of coal-fired generation capacities in the Poland’s market, the impact of coal prices on electricity is high. For East European region, including Poland, the extent to which increasing costs of hard coal are directly passed through to electricity prices at the national power exchanges is the highest in the EU amounting to 70…80% (OECD/IEA 2014). Concerning the impact of electricity prices on hard coal, the IEA assesses high impact, since mid-merit coal plants set the price in 80% of the hours due to renewable expansion and high-cost differences from gas plants (OECD/IEA 2014). Capros (2006) noted that passing through opportunity costs depends on market power in electricity supply. It has nothing to do with the EU ETS system but depends on the ability of competition and regulation to limit potential rents. The full complexity of the pass-through EU ETS costs on electricity prices is studied by Weishaar (2016) who explains differences between theoretical approach and empirical literature.

The second factor affecting energy cost is the cost of EU ETS participation. Pass-through of the costs of decarbonisation on electricity prices addresses the question to what extent the climate policy costs contribute to the rise of energy prices which may result in lower competitiveness and higher energy costs for the end users. EUA low prices of €4.5…6/t CO₂ have not had a notable impact on the electricity market. It is estimated that the increase of the EUA price by one euro results in the wholesale electricity price increase of €0.2…0.8/MWh depending on EU markets (EC 2016d; OECD/IEA 2014). An extensive analysis on the EU ETS pass-through effects on energy prices is presented by Laing et al. (2014). They proved that pass-through cost in electricity can amount to 20…100%. Concerning the EU ETS impact on electricity prices, Lise et al. (2010) showed by simulation that in Poland, wholesale electricity prices are very vulnerable to increase of the EUA price. The pass-through of the EUA cost to electricity price in Poland is extremely high as the portion of allowance price of €20/t CO₂ amounts to €19/t CO₂. Such a high pass-through rate is caused by the coal-dominated power sector. The average for EU-20 examined countries, excluding Poland, was €10…13/MWh. This means 12…27% increase to the wholesale electricity price in the countries due to emission trading. Depending on the fuel prices, mainly the ratio of gas to coal prices, the additional EU ETS-induced price component can change the merit order, e.g. shifting marginal costs (€/MWh) of the Combined Cycle Gas Turbine (CCGT) technology before coal technologies.

Electrical energy prices in Poland have been rather stable over the years considered and do not exhibit a close correlation with EUA prices. Their impact on electricity wholesale prices is not traceable what indicates that power sector still finds a possibility for carbon cost compensation, e.g. in lignite-fuelled power stations being economically merged with lignite mines which give large margins for manoeuvres, or the cost is not yet pass-through to retail prices or is factored in, but at a low and barely visible price. The low electrical energy prices do not either create market signals for technological transformation. The impact is likely to remain modest unless the EUA price rises significantly.