Update docs

docs/transform.rst

@@ -845,11 +845,49 @@ activities to the new regional markets. The regional market it re-links to
 depends on the location of the consumer.
 
 Cement production
-+++++++++++++++++
+"""""""""""""""""
 
-The modelling of future improvements in the cement sector is relatively
-simple at the moment, and does not involve the emergence of new
-technologies (e.g., electric kilns).
+The modelling of future improvements in the cement sector is dependent on the IAM model chosen.
+
+When choosing IMAGE, scenarios include the emergence of a new, more efficient kiln, as well
+as kilns fitted with three types of carbon capture technologies:
+
+* using monoethanolamine (MEA) as a solvent,
+* using oxyfuel combustion,
+* using Direct Separation (Leilac process).
+
+The implementation of the corresponding datasets for these new kiln technologies are based on the work of
+Muller_ et al., 2024.
+
+.. _Muller: https://doi.org/10.1016/j.jclepro.2024.141884
+
+We differ slightly from the implementation of Muller_ et al., 2024, in that:
+
+* the heat necessary for the regeneration of the MEA solvent is assumed to be provided by a natural gas boiler
+(instead of a fuel mix resembling that of the kiln itself), with up to 30% coming from recovered heat from the kiln by 2050,
+* the amount of heat needed for the regeneration of the MEA solvent goes from 3.76 GJ/ton CO2 in 2020, to 2.6 GJ/ton CO2 in 2050,
+* the provision of oxygen for the Direct Separation option comes from an existing air separation dataset from ecoinvent,
+* the fuel mix for the kiln is that of ecoinvent, further scaled down by the change of efficiency of the kiln
+(in Müller et al., 2024, they use directly the fuel mix provided by the IMAGE scenario, which we do not find representative,
+as it also includes the fuel used by other activities in the non-metallic minerals, notably a large share of natural gas).
+
+In a nutshell, *premise*:
+* makes copies of the `clinker production` dataset,
+* adjusts the fuel consumption and related CO2 emissions,
+* adjusts specific hot pollutant emissions removed by the carbon capture process (Mercury, NOx, SOx),
+* adds an input from the carbon capture process, based on a capture efficiency share,
+* and removes a corresponding amount from the outgoing CO2 emissions.
+
+The Direct Separation process only captures calcination emissions, while the other two technologies capture
+both combustion and calcination emissions.
+
+When choosing another IAM, the current implementation is relatively
+simpler at the moment, and does not involve the emergence of new
+technologies. In these scenarios, the production volumes of kilns
+equipped with CCS is not given. Instead, the share of CO2 emissions
+that is sequestered is given. We use the ratio of the CO2 emissions
+sequestered over the total CO2 emissions to determine the share of
+the CO2 emissions that is sequestered in the clinker production dataset
 
 Run
 
@@ -887,17 +925,9 @@ Efficiency adjustment
 ---------------------
 
 *premise* then adjusts the thermal efficiency of the process.
-It does so by calculating the technology-weighted energy requirements
-per ton of clinker.
-Based on GNR/IEA roadmap data, *premise* uses:
 
-* the share of kiln technology for a given region today (2020):
-    * wet,
-    * dry,
-    * dry with pre-heater,
-    * and dry with pre-heater and pre-calciner
-
-* the energy requirement for each of these technologies today (2020).
+It first calculates the energy input in teh current (original) dataset,
+by looking up the fuel inputs and their respective lower heating values.
 
 Once the energy required per ton clinker today (2020) is known, it is
 multiplied by a *scaling factor* that represents a change in efficiency
@@ -916,15 +946,16 @@ between today and the scenario year.
 .. note::
 
     *premise* enforces a lower limit on the fuel consumption per ton of clinker.
-    This limit is set to 2.8 GJ/t clinker and corresponds to the minimum
+    This limit is set to 3.1 GJ/t clinker and is close to the minimum
     theoretical fuel consumption with an moisture content of the raw materials,
-    as considered in the 2018 IEA_ cement roadmap report. Hence, regardless of the
-    scaling factor, the fuel consumption per ton of clinker will never be less than
-    2.8 GJ/t.
+    as considered in the 2018 IEA_ cement roadmap report (i.e., 2.8 GJ/t clinker).
+    Hence, regardless of the scaling factor, the fuel consumption per ton of clinker
+    will never be less than 3.1 GJ/t.
 
 .. _IEA: https://iea.blob.core.windows.net/assets/cbaa3da1-fd61-4c2a-8719-31538f59b54f/TechnologyRoadmapLowCarbonTransitionintheCementIndustry.pdf
 
 
+
 Once the new energy input is determined, *premise* scales down the fuel,
 and the fossil and biogenic CO2 emissions accordingly, based on the Lower Heating Value
 and CO2 emission factors for these fuels.
@@ -943,22 +974,21 @@ If the IAM scenario indicates that a share of the CO2 emissions
 for the cement sector in a given region and year is sequestered and stored,
 *premise* adds CCS to the corresponding clinker production dataset.
 
-The CCS dataset used to that effect is from Meunier_ et al., 2020.
-The dataset described the capture of CO2 from a cement plant.
+The CCS dataset used to that effect is from Muller_ et al., 2024.
+The dataset described the capture of CO2 from a cement plant,
+using a monoethanolamine-based sorbent.
 To that dataset, *premise* adds another dataset that models the storage
 of the CO2 underground, from Volkart_ et al, 2013.
 
 
 Besides electricity, the CCS process requires heat, water and others inputs
 to regenerate the amine-based sorbent. We use two data points to approximate the heat
-requirement: 3.66 MJ/kg CO2 captured in 2020, and 2.6 MJ/kg in 2050.
-The first number is from Meunier_ et al., 2020, while the second number is described
+requirement: 3.76 MJ/kg CO2 captured in 2020 (minus 30% coming from the kiln as recovered heat),
+and 2.6 MJ/kg in 2050. The first number is from Muller_ et al., 2024, while the second number is described
 as the best-performing pilot project today, according to the 2022 review of pilot
 projects by the Global CCS Institute_. It is further assumed that the heat requirement
-is fulfilled to an extent of 15% by the recovery of excess heat, as mentioned in
-the 2018 IEA_ cement roadmap report.
+is fulfilled to an extent of 30% by the recovery of excess heat, as found in numerous studies.
 
-.. _Meunier: https://www.sciencedirect.com/science/article/pii/S0960148119310304
 .. _Volkart: https://doi.org/10.1016/j.ijggc.2013.03.003
 .. _Institute: https://www.globalccsinstitute.com/wp-content/uploads/2022/05/State-of-the-Art-CCS-Technologies-2022.pdf