Ultra-lightweight, high-performance, foldable: advances in space photovoltaics

Daniel Lin­cot. The pho­to­volta­ic effect was dis­cov­ered by Edmond Bec­quer­el in 1839. Sil­i­con solar cells were invent­ed before the Sec­ond World War by Rus­sell Ohl, and the first patent was filed in 1941¹. In 1954, the first effi­cient solar cell was man­u­fac­tured, achiev­ing an effi­cien­cy of 6%. The Unit­ed States, which was look­ing to sup­ply satel­lites with ener­gy, imme­di­ate­ly began pro­duc­ing cells for satel­lites. In 1958, Van­guard 1 was the first satel­lite sent into space with solar cells. The cost of pho­to­voltaics was extreme­ly high at the time [Editor’s note: In the 1950s, one watt of solar pho­to­volta­ic capac­i­ty cost £1,865, adjust­ed for infla­tion and 2019 prices. Com­pared to the price in 1956, a solar mod­ule today would cost £596,800², but afford­able com­pared to the cost of launch­ing a satel­lite into space. The increas­ing pro­duc­tion of solar cells for space appli­ca­tions has led to a decline in costs, which con­tin­ues today. With­out the use of pho­to­voltaics in space, the tech­nol­o­gy would prob­a­bly not have devel­oped as quickly.

Loris Ibar­rart. Satel­lites are autonomous objects, par­tic­u­lar­ly from an ener­gy per­spec­tive. To car­ry out their mis­sions, whether for telecom­mu­ni­ca­tions, mil­i­tary pur­pos­es, or space or Earth obser­va­tion, they must be able to com­mu­ni­cate and to sur­vive. For this they need ener­gy, to send and receive infor­ma­tion from Earth, keep equip­ment at the right tem­per­a­ture and main­tain alti­tude. Final­ly, the mis­sion itself also requires ener­gy. While Earth obser­va­tion mis­sions do not con­sume much ener­gy, telecom­mu­ni­ca­tions do. Ini­tial­ly, a bat­tery was placed on board the satel­lite. Its sur­vival capac­i­ty was only a few weeks. Space indus­try play­ers there­fore thought of using the only resource avail­able in space: the Sun.

DL. The first satel­lite launched into space in 1957, Sput­nik, was only able to com­mu­ni­cate with Earth for a few weeks! It then remained in orbit around Earth, with no means of communication.

LI. From the out­set, there was an idea to use pho­to­volta­ic pan­els on Earth, but the tech­nol­o­gy was only viable for the space industry.

DL. From the 1970s onwards, the first ter­res­tri­al appli­ca­tions emerged: equip­ment for light­hous­es and bea­cons in iso­lat­ed areas, means of com­mu­ni­ca­tion in inac­ces­si­ble loca­tions, and cathod­ic pro­tec­tion for oil pipelines to lim­it oxi­da­tion. These very spe­cif­ic uses jus­ti­fied the high prices. Then the cost of cells began to fall, par­tic­u­lar­ly fol­low­ing the 1973 oil cri­sis, which led to an increase in pro­duc­tion and there­fore a reduc­tion in costs due to economies of scale.

LI. The require­ments of the space indus­try remain the same: to pro­duce as much ener­gy as pos­si­ble on board for the small­est pos­si­ble mass and vol­ume. Most satel­lites orbit­ing the Earth are equipped with pho­to­volta­ic pan­els. Telecom­mu­ni­ca­tions satel­lites have the high­est pro­duc­tion capac­i­ty, which can reach up to around 30 kW, or a solar cell sur­face area of approx­i­mate­ly 100 m². Only a few mis­sions can­not rely entire­ly on solar pow­er: probes sent into deep space and rovers [Editor’s note: “astro­mo­biles”, vehi­cles designed to explore the sur­face of a celes­tial body]. They car­ry a nuclear core, a kind of radioiso­tope bat­tery. This ener­gy source is more expen­sive and restric­tive than photovoltaics.

DL. Pho­to­voltaics have come a long way, par­tic­u­lar­ly in terms of effi­cien­cy, and remain the most effi­cient source of space ener­gy. The effi­cien­cy of the spe­cial cells used (mul­ti-junc­tion) can reach 35%. A record effi­cien­cy of 47% has been achieved in the lab­o­ra­to­ry. In com­par­i­son, the effi­cien­cy of ter­res­tri­al sil­i­con cells is just around 25%.

DL. The the­o­ret­i­cal effi­cien­cy of con­vert­ing pho­tons (light par­ti­cles) into elec­tric­i­ty is 85%. Research is very active in this area, and it is a tremen­dous avenue for progress for ter­res­tri­al appli­ca­tions. Final­ly, space appli­ca­tions increas­ing­ly require very light cells to lim­it launch costs and facil­i­tate deploy­ment. While ter­res­tri­al pho­to­volta­ic pan­els weigh an aver­age of 25 kg per m2, we are work­ing to achieve a weight of 200 g per m2. This is a par­a­digm shift: pho­to­voltaics are becom­ing ultra-light, effi­cient and fold­able. On Earth, this opens up a world of pos­si­bil­i­ties: for exam­ple, we can imag­ine a solar cur­tain that could be deployed on facades, roofs or in the air. These cur­tains could also be tem­porar­i­ly deployed over fields after har­vest to store elec­tric­i­ty. We are work­ing close­ly with the Ile-de-France Pho­to­volta­ic Insti­tute and Ecole Polytechnique’s (IP Paris) inter­face and thin film physics lab­o­ra­to­ry, which spe­cialise in these subjects.

LI. We are also at a turn­ing point for space pho­to­voltaics, with a shift from Earth back to space.

LI. For about 10 years now, we have seen a grow­ing inter­est in ter­res­tri­al tech­nolo­gies in the space indus­try. The rea­son? The rise of satel­lite con­stel­la­tions, for which the space industry’s eco­nom­ic mod­el is no longer com­pat­i­ble. The require­ments here are dif­fer­ent: high pro­duc­tion capac­i­ty, low­er cost and low­er per­for­mance, since redun­dan­cy is ensured by the large num­ber of satel­lites. Today, the cost of cells for space appli­ca­tions is around €300 per watt, com­pared with 10–20 cents for ter­res­tri­al appli­ca­tions. We can imag­ine a com­pro­mise emerg­ing between mass pro­duc­tion for ter­res­tri­al appli­ca­tions and pre­ci­sion work for space appli­ca­tions, even if the future is uncer­tain. The chal­lenge is to mod­i­fy ter­res­tri­al cells so that they can oper­ate for as long as pos­si­ble in space, while mak­ing as few changes as pos­si­ble to exist­ing indus­tri­al chains.

LI. No, the rise of satel­lite con­stel­la­tions will not kill off the tra­di­tion­al space solar pan­el indus­tries. Telecom­mu­ni­ca­tions, obser­va­tion and sci­ence will con­tin­ue to need process­es devel­oped specif­i­cal­ly for space. Man­u­fac­tur­ers have nev­er seen as much demand as they do today.

Interview by Anaïs Marechal