ZeroCO2Glas: Development of a novel hydrogen-fired glass melting furnace with the aim of a CO₂-neutral container glass production

The over­all goal of the ZeroCO2Glas pro­ject is to deve­lop a revo­lu­tio­na­ry glass mel­ting pro­cess in con­nec­tion with a new type of glass mel­ting tank in an all-encom­pas­sing approach, with which con­tai­ner glass can be mel­ted in a CO₂-neu­tral man­ner and with signi­fi­cant ener­gy savings. The dura­ti­on of the pro­ject is three years and it is fun­ded by the BMWK. The fol­lo­wing pro­ject part­ners are invol­ved in the rese­arch project:

Over­view of the pro­ject consortium

In order to achie­ve the cli­ma­te tar­gets, mas­si­ve emis­si­on reduc­tions must also be rea­li­sed in the ener­gy-inten­si­ve glass indus­try. The pro­ject focu­ses on the con­tai­ner glass indus­try, which accounts for the lar­gest share of glass pro­duc­tion in Ger­ma­ny at just over 50%. Depen­ding on the type of glass and the pro­duc­tion pro­cess, up to 85% of the ener­gy requi­red in the manu­fac­tu­ring pro­cess is nee­ded for the mel­ting pro­cess: here, the mix­tu­re of and was­te glass cul­let must be hea­ted to a tem­pe­ra­tu­re of 1450 °C to 1650 °C and con­ver­ted into glass. The glass fur­naces are usual­ly fired with natu­ral gas. In lar­ge mel­ting tanks, this is main­ly done by natu­ral gas-fired bur­ners. In addi­ti­on, the fur­naces can be equip­ped with an elec­tric auxi­lia­ry hea­ter: In this case, up to 20% of the mel­ting power is intro­du­ced by elec­tro­des that usual­ly pro­tru­de ver­ti­cal­ly into the mol­ten glass (see illustration).

Cross fired hybrid mel­ting fur­nace [Horn]

Within the scope of the pro­ject, a new, inno­va­ti­ve mel­ting tech­no­lo­gy is to be deve­lo­ped and tes­ted. The new glass mel­ting tank should requi­re 15% less ener­gy than con­ven­tio­nal mel­ting fur­naces. This is achie­ved by using an alter­na­ti­ve, car­bo­na­te-free batch with a lower mel­ting enthal­py to avo­id raw mate­ri­al-rela­ted CO₂ emis­si­ons. On the other hand, the ener­gy requi­re­ment is redu­ced due to the dis­pensable mois­tening of the feedstock and reduc­tion of the resi­dence time. The ener­gy-rela­ted CO₂ emis­si­ons are also to be eli­mi­na­ted by swit­ching to hydro­gen-oxy­fuel combustion.

At the Depart­ment for Indus­tri­al Fur­naces and Heat Engi­nee­ring (IOB), expe­ri­men­tal inves­ti­ga­ti­ons and nume­ri­cal simu­la­ti­ons of the com­bus­ti­on space in the upper fur­nace and of the glass melt flow in the lower fur­nace are being car­ri­ed out as part of the pro­ject. The aim of work packa­ges 4 and 5 is to build an accu­ra­te simu­la­ti­on model and to inves­ti­ga­te the influen­ces of fur­ther para­me­ters in order to final­ly deve­lop a best-case sce­na­rio for the pilot fur­nace.
For the CFD simu­la­ti­on of the com­bus­ti­on, estab­lished models are the­r­e­fo­re first che­cked for their appli­ca­bi­li­ty to hydro­gen-oxy­fuel com­bus­ti­on by means of vali­da­ti­on on a bur­ner test bench and adapt­ed if neces­sa­ry. This is fol­lo­wed by the simu­la­ti­on of the pilot tank. After suc­cessful vali­da­ti­on mea­su­re­ments at the pilot plant, the CFD model is then used to inves­ti­ga­te the influen­ces of fur­ther para­me­ters such as the bur­ner incli­na­ti­on ang­le or the power dis­tri­bu­ti­on on the mel­ting pro­cess and ener­gy effi­ci­en­cy.
The influen­ces of the increased pro­por­ti­on of elec­tri­cal mel­ting power and of the sub­mer­ged fee­der on the flow within the glass melt will be inves­ti­ga­ted both phy­si­cal­ly and nume­ri­cal­ly. For this pur­po­se, a phy­si­cal acrylic glass model of the pilot tank is first desi­gned using the simi­la­ri­ty theo­ry. In this model, the flow of a model flu­id simi­lar to the glass melt is inves­ti­ga­ted using par­tic­le image velo­ci­me­try (PIV). Fur­ther­mo­re, a CFD model of the acrylic glass test rig is built and vali­da­ted with the expe­ri­men­tal results. The fin­dings from the phy­si­cal and nume­ri­cal simu­la­ti­on of the model flu­id are then incor­po­ra­ted into the CFD simu­la­ti­on of the glass melt. The second nume­ri­cal model maps the pilot tank’s sub-fur­nace and enables the flow inves­ti­ga­ti­on within the glass melt: Among other things, the influence of the sub­mer­ged fee­der com­pared to the con­ven­tio­nal batch fee­der on the melt pool is deter­mi­ned, but also the power dis­tri­bu­ti­on of the elec­tro­des is opti­mi­sed with regard to impro­ved ther­mal­ly indu­ced flow control.