Christophe Forgez and Nico­las Damay, tenured pro­fes­sor and lec­tur­er, respec­tive­ly, are research sci­en­tists at the Rober­val Lab­o­ra­to­ry. Spe­cial­is­ing in mod­el­ling, they work on projects relat­ed to sodi­um-ion bat­ter­ies. This inno­v­a­tive tech­nol­o­gy com­bines, among oth­er things, a con­cern for sov­er­eign­ty and a con­cern for envi­ron­men­tal impact.

When he first arrived at UTC in 1999 as a lec­tur­er, Christophe Forgez was attached to the Com­piègne Electro­mechan­ics Lab­o­ra­to­ry (LEC), which has since merged with UTC-Rober­val. This lab­o­ra­to­ry spe­cialis­es in embed­ded elec­tri­cal ener­gy sys­tems with two areas of exper­tise: elec­tri­cal machines and pow­er con­vert­ers. How­ev­er, it lacked the bat­tery com­po­nent as an ener­gy source.

This led him to become inter­est­ed in var­i­ous issues relat­ed to bat­tery tech­nolo­gies in the ear­ly 2000s, par­tic­u­lar­ly mod­el­ling for inte­gra­tion into pow­er­train con­trol. His research led him to forge aca­d­e­m­ic and indus­tri­al part­ner­ships at a time when elec­tric vehi­cles were not yet on the hori­zon, but hybrid vehi­cles were. ‘That’s how, as part of the joint lab­o­ra­to­ry with Valéo, we start­ed work­ing on ‘mild hybrid’ sys­tems, which were essen­tial­ly the first steps towards hybridi­s­a­tion,’ he says.

Christophe Forgez and his team then nat­u­ral­ly moved on to eval­u­at­ing the per­for­mance of dif­fer­ent bat­tery mod­els, as these had to meet the require­ments of hybrid appli­ca­tions. ‘We estab­lished an ini­tial col­lab­o­ra­tion with the LRCS, an elec­tro­chem­istry lab­o­ra­to­ry in Amiens head­ed at the time by Pro­fes­sor Taras­con, as part of the region­al DIVA (Advanced Vehi­cle Diag­nos­tics) project, which was ded­i­cat­ed to bat­tery mod­el­ling. Our aim in this col­lab­o­ra­tion was to under­stand the elec­tro­chem­i­cal phe­nom­e­na occur­ring at the heart of the bat­tery so that we could trans­late them using our own mod­el­ling tools. We were thus able to set up an ini­tial mod­el struc­ture with localised con­stants using equiv­a­lent elec­tri­cal dia­grams of the inter­nal elec­tro­chem­i­cal phe­nom­e­na,’ he explains.

New skills

As the lab­o­ra­to­ry wel­comed new skills, new issues relat­ed to bat­ter­ies were explored. They began work­ing on esti­mat­ing inter­nal quan­ti­ties such as state of charge, for exam­ple. This is an area of inter­est to man­u­fac­tur­ers, par­tic­u­lar­ly car man­u­fac­tur­ers. ‘In a bat­tery, you can mea­sure the volt­age or the incom­ing cur­rent, but there is no sen­sor to assess its state of charge. As part of a the­sis, we have there­fore devel­oped a mod­el that reports the state of charge with a high degree of accu­ra­cy. This is known as a “state of charge observ­er”,’ says Christophe Forgez.

With the proof of con­cept for the state of charge observ­er hav­ing been estab­lished in the first the­sis, a sec­ond the­sis has been launched in part­ner­ship with Renault to improve state of charge esti­ma­tors. “The first elec­tric cars had sig­nif­i­cant ener­gy reserves hid­den from the user to avoid range anx­i­ety. In fact, vehi­cle man­u­fac­tur­ers were seek­ing to instal (and did so) an ener­gy reserve, only part of which was used. This pre­cau­tion required more bat­ter­ies, which meant more weight and addi­tion­al cost. The esti­ma­tors we pro­posed, inte­grat­ed into the BMS (bat­tery man­age­ment sys­tem), had to guar­an­tee the end of dis­charge with a high degree of accu­ra­cy. This enabled the man­u­fac­tur­er to reduce the weight of the vehi­cle and increase its range. How­ev­er, it is impor­tant to note that a bat­tery’s capac­i­ty decreas­es as the vehi­cle and its bat­ter­ies age. The para­me­ters there­fore need to be read­just­ed in real time,’ he explains.

But bat­tery tech­nol­o­gy is mul­ti­phys­i­cal and requires cou­pling between elec­tro­chem­i­cal and ther­mal phe­nom­e­na. This led to a part­ner­ship with E4V (Ener­gy for Vehi­cles), a design­er and man­u­fac­tur­er of bat­ter­ies for mobil­i­ty appli­ca­tions. The com­pa­ny was acquired in 2024 by Arts Ener­gy, a major play­er in elec­tri­cal ener­gy stor­age. ‘The aim was to deter­mine a ther­mal mod­el for the bat­tery, giv­en that it heats up dur­ing charg­ing and dis­charg­ing. What’s more, the more it heats up, the faster it ages. E4V want­ed to make sure that the bat­tery packs it man­u­fac­tures could with­stand cer­tain appli­ca­tions. What is the opti­mum heat lev­el? Did the cool­ing sys­tem need to be redesigned or the num­ber of cells increased, for exam­ple? This was the prob­lem Nico­las Damay was asked to solve by the com­pa­ny,’ he explains.

Battery ageing

Nico­las Damay worked part-time with E4V dur­ing his the­sis. ’Using mod­els, we tried to pre­dict how tem­per­a­tures would change. Tem­per­a­ture had been iden­ti­fied in a num­ber of nation­al projects in which UTC was involved as one of the main fac­tors in bat­tery age­ing. How­ev­er, in order to analyse the ther­mal prop­er­ties prop­er­ly, it was also nec­es­sary to study the elec­tro­chem­i­cal prop­er­ties and devel­op mod­els that would be reli­able through­out the bat­tery’s ser­vice life,’ he believes.

Recog­nis­ing the inter­con­nec­tion between elec­tro­chem­i­cal and ther­mal phe­nom­e­na, Nico­las Damay decid­ed to take things fur­ther dur­ing his post-doc­tor­ate and lat­er as a senior lec­tur­er. ‘My idea was that if we could use the mod­els we had devel­oped to deter­mine the state of charge of a bat­tery, for exam­ple, we should also be able to analyse oth­er more detailed com­po­nents. For exam­ple, if we look at age­ing, we see that degra­da­tion begins right from the start of the bat­tery’s life. Close­ly mon­i­tor­ing this phe­nom­e­non is not easy, as they are local. Detect­ing what is hap­pen­ing at the elec­trodes, for exam­ple, with­out open­ing the bat­tery posed a num­ber of chal­lenges in my view,’ he explains.

What are the caus­es of age­ing? ‘It is the growth of a pas­si­va­tion lay­er on the neg­a­tive elec­trode (SEI: sol­id elec­trolyte inter­phase), main­ly due to high tem­per­a­tures, that is respon­si­ble for the loss of bat­tery auton­o­my in the first few years. How can we detect the impact of SEI and accu­rate­ly pre­dict its evo­lu­tion by mea­sur­ing the out­put volt­age? This is how we devel­oped a mod­el that incor­po­rates more physics and is capa­ble of detect­ing the evo­lu­tion of this phe­nom­e­non,’ he says.

A variety of technologies

This research issue has led to sev­er­al pub­li­ca­tions. “We also had to ver­i­fy that the mod­el adapts to dif­fer­ent chemistries. Lithi­u­mion, for exam­ple, is a fam­i­ly of chem­i­cal­ly very dif­fer­ent objects. While lithi­um ions flow from one elec­trode to anoth­er, the elec­trode mate­ri­als that store these ions dif­fer depend­ing on the mod­el of bat­tery. These mate­ri­als can be based on man­ganese, nick­el or cobalt, which are very effi­cient but pose prob­lems in terms of both abun­dance on Earth and sup­ply. They can also be made from iron phos­phate, which is more abun­dant. Lithi­um ions can also be replaced by sodi­um ions. Tia­mat, a spin-off of the LRCS in Amiens, has devel­oped a bat­tery based on sodi­um-ion tech­nol­o­gy with the fol­low­ing dis­tinc­tive fea­ture: the mate­ri­als used are also more abun­dant and more acces­si­ble in Europe. Cur­rent­ly, the per­for­mance of this bat­tery is slight­ly low­er than that of lithi­u­m­iron- phos­phate (LFP) bat­ter­ies used in elec­tric vehi­cles, but it is already emerg­ing as a com­peti­tor. A gigafac­to­ry is cur­rent­ly under con­struc­tion near Amiens in part­ner­ship with Stel­lan­tis, with a pro­duc­tion line ded­i­cat­ed to Tia­mat bat­ter­ies. The first phase is sched­uled for 2027,” he adds.

What are the spe­cif­ic fea­tures of bat­tery research at UTC? “Else­where, chem­istry lab­o­ra­to­ries focus on lab­o­ra­to­ry bat­ter­ies for devel­op­ing mate­ri­als and man­u­fac­tur­ing process­es. At UTC, we use the knowl­edge accu­mu­lat­ed by elec­tro­chemists to improve the diag­no­sis and prog­no­sis of com­mer­cialscale objects. This allows us to estab­lish links between more fun­da­men­tal knowl­edge and the needs of inte­gra­tors,’ says Nico­las Damay.

A vari­ety of tech­nolo­gies is used in the man­u­fac­ture of bat­ter­ies for elec­tric vehi­cles. “Today, more and more elec­tric vehi­cles are run­ning on LFP bat­ter­ies, a less expen­sive tech­nol­o­gy than oth­er lithi­um-ion tech­nolo­gies. LFP requires less spe­cial care in the man­u­fac­tur­ing process, mak­ing it less expen­sive to pro­duce. Chi­na is the leader in this field. There is also NMC (nick­el-man­ganese-cobalt) tech­nol­o­gy, which offers very high per­for­mance, but these mate­ri­als are very crit­i­cal in terms of abun­dance, geo­graph­i­cal loca­tion and there­fore sov­er­eign­ty. In the future, with greater indus­tri­al matu­ri­ty, Tia­mat’s sodi­u­mion tech­nol­o­gy could offer a real alter­na­tive, par­tic­u­lar­ly in the seg­ment cur­rent­ly occu­pied by LFP tech­nolo­gies. Tia­mat’s tech­nol­o­gy has sev­er­al advan­tages. First, it uses sodi­um, which is found every­where, which is sig­nif­i­cant in terms of both sov­er­eign­ty and envi­ron­men­tal impact. In addi­tion, it per­forms bet­ter in fast charg­ing and final­ly lasts longer, with more than 10 000 charge-dis­charge cycles, com­pared to a few thou­sand for most lithi­u­mion tech­nolo­gies,” explains Nico­las Damay

In the aero­space indus­try, a car­bon-free tar­get has been set for 2040. “To meet this tar­get, tur­bine hybridi­s­a­tion projects are being stud­ied, as well as all-elec­tric engines (Air­bus’ E‑Fan, VTOL for Ver­ti­cal Take-Off & Land­ing, etc.). The aero­space sec­tor is there­fore increas­ing­ly inter­est­ed in lithi­um-ion tech­nolo­gies as an on-board ener­gy source. It is in this con­text that we have been work­ing with Safran since 2013. Rail trans­porta­tion is anoth­er sec­tor that is inter­est­ed in new bat­tery tech­nolo­gies for its loco­mo­tives and the rail­road net­work, and we are also work­ing with this sec­tor,’ says Christophe Forgez.

The challenge of fast charging

With the rise of elec­tro­mo­bil­i­ty, a race to devel­op fast charg­ing has begun. This is an area in which UTC is heav­i­ly involved. “In our work, we are able to char­ac­terise bat­ter­ies from a ther­mal and elec­tro­chem­i­cal point of view, but we also have suf­fi­cient exper­tise to define their per­for­mance lim­its. We are cur­rent­ly work­ing on the issues of fast and even ultra-fast charg­ing. This is an area of inter­est to many man­u­fac­tur­ers in the elec­tric mobil­i­ty sec­tor, as well as to elec­tric­i­ty net­works. As part of a the­sis, we have man­aged to charge a cell by 30% in two min­utes at 20°C,” says Christophe Forgez.

This is a chal­lenge for many play­ers keen to offer ever more ser­vices to users. “We are work­ing on the sodi­um-ion tech­nol­o­gy, which has anoth­er advan­tage: it allows faster charg­ing. In fact, we could end up with a mar­ket where vehi­cles with a long range but which can­not be charged quick­ly coex­ist with vehi­cles with a short­er range but which, con­verse­ly, can be charged in a few min­utes. This would allow for the deploy­ment of dif­fer­ent types of charg­ers. In super­mar­ket car parks, for exam­ple, they would not need to be fast charg­ers as cus­tomers tend to stay for rel­a­tive­ly long peri­ods of time. Con­verse­ly, on motor­way ser­vice sta­tions, there should be both, as there will be motorists who stop to rest and are hap­py with a slow­er charge, but also those who stop just to charge their vehi­cle and will there­fore need a fast charge,’ says Nico­las Damay.

MSD