element
In regolith ppm
In iron ppm
In iron %

Vanadium
114
863
0.09%

Chromium
2640
20000
2.00%

Manganese
1740
13180
1.30%

Nickel
169
1280
0.12%

Cobalt
40.3
305
0.03%
Speculations: If all the iron and magnetic alloying elements are pulled out, and iron is at 13.2%, then concentrations increase 7.576 times. This is what we might have in the steel. When we figure out how much carbon to allow and how best to heat treat the steel, we will have excellent steel. Eventually we will experiment with REE additions and even extract trace elements to control composition of the steel. We might even alloy with zirconium if we can find deposits or extract traces. Steel can be alloyed with titanium and silicon too.

We will build more molten silicate electrolysis units on the Moon from sintered basalt and ceramic blocks from electrolysis, steel or titanium, piled regolith for thermal insulations, rutile lined titanium oxygen outlet pipes, rutile lined steel or titanium metal and ceramic taps, etc. What will we use for electrodes? Irridium? No. FeSi? It melts at about 1200 C. and the furnace will run at 1300 to 1600 C. However, it has been suggested the ferrosilicon be used for electrodes as they will be in thermal balance with the melt.  Perhaps fused CaO will make workable electrodes. This ceramic can be heated to incandescence, the limelight. It melts at 2570 C. From Lime and Magnesia, 1924, E. Benn, we find that the electrical resistance of CaO decreses with temperature. Or will it dissolve in the corrosive melt? 
With enough energy to deoxidize and extract silicon, the manganese, chromium, sodium, phosphorus, potasssium and IRON will all be extracted. The Mn and Cr will be right where we want it-in the iron if we chose to carbruize iron from the magma furnace. The sodium and postassium will boil out. Na b.p. 890 C. and K b.p. 754 C. The furnace runs at 1200 C. or higher. If the voltage is high enough the magnesium will boil out and react with oxygen according to Larry Haskin in a paper on the web I can no longer find and did not download. Mg vapors and O2 could ignite. It would take a lot of electricity to extract the Mg this way and electricity generation is not as efficient as direct solar energy and thermal energy from a power tower, so I stick with other ways. The furnace would have to take a lot of heat also. Though we might try this for a while. If small amounts of Na and K vapor come off and react with oxygen to form Na2O and K2O they might condense and settle back into the molten silicates. Na2O decomposes at 1132 C. and K2O at 350 C. So they would be decomposing and burning with O2 again in some state of equilibrium, depending on vapor temperatures. Now, if we flush the furnace with an inert gas and propell the Na, K, Na2O and K2O out with the oxygen and condense them we have a way to get lots of these substances that are needed for 1) molten salts for the power tower 2) molten salt pots for heat treating steel 3) life support, table salt, fertilizer, even lye for soap making! 4) breeder reactors? 5) sodium vapor lamps? 6) soda-lime glass making 7) any other uses for sodium and potassium? Phosphorus also has a low b.p. And is combustible in oxygen. Haven't given it as much thought but we could flush it through and condense it also. The sodium, potassium, their oxides, phophorus and oxygen might react to form phosphate salts that we flush the tiny particles of thru with inert gas. Upported argon or nitrogen or even helium 4 harvested on the Moon. Helium is leaky but useful stuff. As for separating the Na, K and P that might be done by distillation.

Sodium and potassium and/or their oxides will flush thru with the oxygen. The inert gas will separate from oxygen during liquifaction and we could store it in high pressure tanks since argon and helium liquefy at such low temps. Nitrogen liquefys at about the same temp. as oxygen. Lunar helium anybody?
More Information 4
                                                  About Solar Thermal Power

Solar thermal can be had with troughs or power towers or dishes. The power tower looks best to me and we can get the salt. The best thing to do is use the heat for processes. Heating magnesium retorts because sun light might be reflected to much by the slag in the charge and the window might foul with magnesium vapors condensing. It can preheat gases going into blast furnaces and open hearth furnaces to 1000 C. and higher. It can heat sulfuric acid and dry the silica+CaSO4 mixture as well as the Al and Mg sulfate mixture for full recovery of H2O and acid. Then we can just dump the dried stuff into skip cars in the vacuum. Heat from the power tower can be stored in molten salts for heat and electrical power at night on the Moon. The trough collectors don't get hot enough for any of this nor can they get salts hot enough for good power storage. In the original "Damascus Project" I suggested preheating gases with troughs- bad idea. Power tower-good idea.

As for turbogenerators these are complex devices. We could make iron armatures with aluminum windings and aluminum stator windings for AC generators. We could make iron or steel casings for AC generators. Turbines are another story. Do we need special steels? Do we run them on steam, or do we use hot high pressure carbon dioxide which will not corrode or scale the blades as water will. CO2 is fairly dense and can take hi temps. It's fairly inert up to about 1000 C.

Another problem is heat rejection from turbines. Radiators could do the job. We can't reject heat into the regolith because it is a damn good thermal insulator. But what about the deeper packed down hard regolith? About what, ten feet down when astronauts could not drill any deeper? Is the deep, hard packed, water less, regolith a good heat conductor? Could we dig into it, drill holes, insert heat rejection loops and rejection heat into the ground even during the scorching lunar day because this ground is minus 4 C. ( minus 20F.) all the time covered by many feet of insulating loose packed regolith?

Meanwhile, can we make turbines? What about something made of steel almost as simple as a water wheel in a casing to turn generators? Might not be the most efficient but it might give us enough AC power for arc welders, motors (for compressors, coolers using CO2 refrigerant, pulverizers, etc) and heat probes. Or must we invert D.C. From solar panels so we can step it up to high voltages with transformers???

He who can build a turbine on the Moon made of lunar materials is a genius in my humble opinion. And so is anybody who can work out structural and thermal stresses on things so we can make them last and not crack, collapse, melt, etc.
A power tower can supply process heat to magnesium refining. This might work better than direct application of solar because slag in retort might reflect sun light. A power tower can provide temps. Of 550 to 1500 C. Silicothermic magnesium reduction requires 1550 to 1600 C. NaCl melts at 801C and boils at 1413 C. KCl mp 770 C bp 1550 C. With a power input of 4.5 MW the Magnatherm process can produce 12 tons of magnesium per 24 hours. 9MWhours per ton Mg.

Solar II will be 300 ft tall, have 1,926 sun tracking mirrors, and generate 10 MWe, so it might actually produce 30-40 MW thermal. We could scale things down in the early stages of lunar industrialization. We don't need 12 tons of Mg a day or 30-40 MWt at first.