If by "solids content" you mean the TDS, then the easiest way would be to measure conductivity, and then use an empirical relationship (for example from other test work of the same materials, say in humidity-cell tests, or measured seepage) to correlate TDS to conductivity. A multi-attribute meter, reading say conductivity, pH and ORP, might provide you value additional information to support your interpretation.

From some empirical analysis, I've correlated TDS with conductivity as EC=0.6*TDS, but in this case I need to measure the suspended solids concentration of a sludge bleed, or the pulp solids concentration.

I was afraid you meant suspended solids. Well, there are turbidity meters - Mettle and YSI make them; probably others, too. As with conductivity/TDS, you would need to establish an empirical relationship, and I don't know how reproducible that would be. You could get a meter and make some measurements to see. Or perhaps work with a lab you do work with often to use their meter and see what you can determine about applicability.

Do you mean 'on line' or in a laboratory? What is the solid?

Thanks a lot for your reply. Well, I need to determine the solids concentration of the slurry of the Sludges bleed from the AMD treatment plant. I mean, I need to calculate the solids percentage in the discharge.

There are devices out there that claim to measure solids concentration by usually they are not useful as there are too many factors to take into consideration. On a practical sense, so long as you are talking about actual solids (i.e. metals) and not dissolved chemicals such as lime, the approach we take is to develop a model based on actual lab analysis between solids concentration and liquid SG. In line monitors then measure SG. This is converted back into solids concentration.

The easy was is gravimetrically but for this you would require about a liter of the bleed solution to get a meaningful result for what I assume is a very dilute suspended solid.

Have you got the gravimetric procedure?

I would appreciate your help.

There are many documented methods like the EPA method at the web address below, or "Standard Methods". Generally for dilute solutions you will want 1 L of solution but the linked method indicates 100 ml. The general method is simple. You take a 0.45 micron membrane filter; pass a known volume of solution through the membrane by vacuum filtration, ensuring that any solids are kept in solution by periodic agitation to ensure they are collected on the membrane. Once complete, dry the membrane and determine the increase in mass in mg. Divide the result by the volume passed through the filter to determine total suspended solids in mg/L

http://www.epa.gov/region9/qa/pdfs/160_2.pdf

Correlating suspended solids and turbidity is possible as long as the SS content is low, but the correlation is specific to each fluid. So you may try and calibrate the SS contents vs. turbidity by measuring SS by filtration on different samples of the same slurry. You will need good turbidity standards up to 1000 NTU. Better use readymade standards rather than prepare them yourself unless you are a chemist.

Perhaps you can clarify again? In the previous comment, I understood you initially to be asking for a system that would allow you to track TSS *in-situ*, for "real-time" QC. This of course is perfectly compatible with calibration, which certainly is essential, but your routine measurement is with the remote instrument, such as a turbid meter.

Another specific choice you will have to make is the filter mesh. 0.45 µm is inherited from drinking water bacteriology, but has the advantage to be known by everyone and be in plentiful supply.

If you are dealing with heavily loaded slurries, you may prefer using a 5 µm mesh.

But 0.45 µm is not enough if you plan to separate TDS from TSS. Lots of particles are smaller. Filtering at 0.1 or 0.2 µm can be an option if you want as much as possible TDS in the filtrate.

However, you will still miss some colloids (ultra filtration required... and tedious). But I doubt this is really an issue for AMD or ARD, unless you are in advanced modelling.