15 Methodology for Modeling the Energy …
227
(2016), Lange (2015), Andrae and Edler (2015). Furthermore, despite some product
lifecycle assessments, there are even less studies that investigate and quantify the
overall material demand of ICT and mobile communication networks in particular.
The existing studies on ICT energy consumption are mostly top-down energy
intensity estimates or bottom-up power consumption calculations based on publicly
available product stock data (Andrae 2019). These studies differ in product and
geographic scope. Their results are difficult to compare. In terms of methodology,
these studies often allocate an average power consumption to a representative base
case product and multiply this energy with the product stock. There are considerable limitations to the results of this principle approach. The granularity of the
product selection for instance determines the accuracy of the result. The active mode
power consumption of some product categories such as computers could varies by a
view hundred watts depending on the CPU and RAM configuration. Furthermore, the
studies are assuming, for the sake of simplicity, usually no real load patterns or differentiated usage patterns. Therefore, increasingly higher utilization and respective
power consumption is not adequately reflected.
As a first conclusion, it is justified to say that the existing studies on overall energy
consumption of ICT are mostly broad simplifications of reality. Although the older
modeling approaches have shown plausible realistic results, they are less suitable
as an analytic tool. An advanced inventory model seems necessary for creating and
analyzing precise technology- and usage-dependent scenarios.
With regard to the optimization of the energy efficiency of mobile networks, especially with regard to the concrete 5G development, there are a number of recent studies
with power consumption models [see references Fiorani (2016), Xiaohu (2017),
Nasim (2017), Poirot (2017). A basic power consumption model has been developed in conjunction with the European Union ICT-EARTH project (Imran et al.
2011). This model specifies the main energy contributing components such as power
amplifier, baseband unit, cooling, and power supply losses. This power model is quit
abstract and not granular enough to calculated energy efficiency options. More recent
studies present detailed models that allow, for example, comparison of transmission
power with computation power in different network scenarios (Fiorani 2016; Xiaohu
2017). However, these models are also focused only on the analysis of individual
technology. None of the models has the goal of calculating a real network implementation. Furthermore, only the energy consumption is modeled. The specific material
requirements are not the subject of these studies.
There is another aspect that needs consideration. Past studies did not put the energy
and material consumption in perspective to the relative functional performance. The
energy performance of a network for instance is a ratio between the amounts of
energy used for the transport of a specific amount of data. It makes sense to show
the correlation between energy consumption and performance in order to assess the
relative environmental impact of telecommunication networks.
227
(2016), Lange (2015), Andrae and Edler (2015). Furthermore, despite some product
lifecycle assessments, there are even less studies that investigate and quantify the
overall material demand of ICT and mobile communication networks in particular.
The existing studies on ICT energy consumption are mostly top-down energy
intensity estimates or bottom-up power consumption calculations based on publicly
available product stock data (Andrae 2019). These studies differ in product and
geographic scope. Their results are difficult to compare. In terms of methodology,
these studies often allocate an average power consumption to a representative base
case product and multiply this energy with the product stock. There are considerable limitations to the results of this principle approach. The granularity of the
product selection for instance determines the accuracy of the result. The active mode
power consumption of some product categories such as computers could varies by a
view hundred watts depending on the CPU and RAM configuration. Furthermore, the
studies are assuming, for the sake of simplicity, usually no real load patterns or differentiated usage patterns. Therefore, increasingly higher utilization and respective
power consumption is not adequately reflected.
As a first conclusion, it is justified to say that the existing studies on overall energy
consumption of ICT are mostly broad simplifications of reality. Although the older
modeling approaches have shown plausible realistic results, they are less suitable
as an analytic tool. An advanced inventory model seems necessary for creating and
analyzing precise technology- and usage-dependent scenarios.
With regard to the optimization of the energy efficiency of mobile networks, especially with regard to the concrete 5G development, there are a number of recent studies
with power consumption models [see references Fiorani (2016), Xiaohu (2017),
Nasim (2017), Poirot (2017). A basic power consumption model has been developed in conjunction with the European Union ICT-EARTH project (Imran et al.
2011). This model specifies the main energy contributing components such as power
amplifier, baseband unit, cooling, and power supply losses. This power model is quit
abstract and not granular enough to calculated energy efficiency options. More recent
studies present detailed models that allow, for example, comparison of transmission
power with computation power in different network scenarios (Fiorani 2016; Xiaohu
2017). However, these models are also focused only on the analysis of individual
technology. None of the models has the goal of calculating a real network implementation. Furthermore, only the energy consumption is modeled. The specific material
requirements are not the subject of these studies.
There is another aspect that needs consideration. Past studies did not put the energy
and material consumption in perspective to the relative functional performance. The
energy performance of a network for instance is a ratio between the amounts of
energy used for the transport of a specific amount of data. It makes sense to show
the correlation between energy consumption and performance in order to assess the
relative environmental impact of telecommunication networks.