Genetic Mutations

I am working on a theory so i can calculate the amount of nutrients in my compost based on the input materials.. in my app there are inputs that i can adjust the input materials, and with them the app derives the amount of elements going in to it and outputs the difference of ( nutes provided - nuets requd ).

attached is an AI to adjust the amount of input material and calcs fitness where 0 is the target like this:

internal decimal checkFitness(string p1, string p2, string p3, string p4, string p5, string p6)
{
decimal s1 = decimal.Parse(p1) - TargetN; if (s1 < 0) { s1 *= -1; }
decimal s2 = decimal.Parse(p2) - TargetP; if (s2 < 0) { s2 *= -1; }
decimal s3 = decimal.Parse(p3) - TargetK; if (s3 < 0) { s3 *= -1; }
decimal s4 = decimal.Parse(p4) - TargetC; if (s4 < 0) { s4 *= -1; }
decimal s5 = decimal.Parse(p5) - TargetM; if (s5 < 0) { s5 *= -1; }
decimal s6 = decimal.Parse(p6) - TargetS; if (s6 < 0) { s6 *= -1; }
return( s1 + s2 + s3 + s4 + s5 + s6);
}


Here is how im mutating my offspring right now:

offspring = new offspringItem(nuteList.Count);
int index = 0;
foreach (nuteObject o in nuteList)
{
decimal rnd22 = 0;
decimal rnd23 = 0;
int rnd1 = random.Next(0, offspringList.Count/2); // select random offspring at top of list
rnd22 = random.Next(-200, +200); // get random
rnd22 /= 100; // convert random to 2 decimal places
rnd23 = offspringList[rnd1].arr4[index] + rnd22; // add random to random selections (i thought multiply would be better but it wasnt)
if (rnd23 > 99) { rnd23 = 99; } // keep within program scope
if (rnd23 <= 0) { rnd23 = 0.02M; } // keep within program scope
NumericList[index].Value = rnd23; // setup in app
offspring.import(index, rnd23); // save new child parameters
index++;
}


however the GA stops finding better solutions at a fitness of about 7-10 , maybe there is a suggestion to generate healthier offspring??