This question is lacking a couple of key pieces of information about the solids settling problem, but I can use common information to get a good estimate of the potential problems.  The density of the solids (500 kg/m^3) is obviously a bulk density, not a particle density.  If the particle density were 500 kg/m^3, they would float on the water at 1,000 kg/m^3.  Floating solids are a more difficult problem than settling solids.
 
To accurately assess solids suspension, a particle size, or size distribution, along with a particle (material) density is necessary to predict how rapidly solids will settle.  However, based on your industry, the solids (particle) density is not high, like mineral ore, and the size is probably less than a few hundred microns.  Using some reasonable estimates for the missing information, the solids suspension capabilities of the mixer should be fairly good, although several other reasons exist for why you have problems.
 
Without questions the impeller is located very near the bottom, probably to assure mixing suspension at low liquid levels.  This bottom impeller will always do most of the actual solids suspension.  Adding an upper impeller will do little to lift more solids off the bottom, it will only reduce the solids concentration near the bottom.  Concentration does not have a strong effect on solids suspension, unless it is high enough to change the apparent fluid viscosity.
 
The real problem may be with the concentration of solids and their effect on the fluid properties.  Whether you are dealing with a 1 to 1.5 or a 1 to 2 concentration and whether that ratio is based on volume or density, you have a high concentration of solids.  Taking the more dilute ratio, you could be talking about 1 volume (m^3) of solids (500 kg) in 2 volumes (m^3) of water (2,000 kg), which gives you 500 kg solids in 2,500 kg of mixture for 20 wt%.  If you are talking masses, you could have 1,000 kg of solids in 2,000 kg of water, or a 33 wt% mixture.  The higher ratio 1 to 1.5 only increases these concentration estimates.
 
Any concentration above 20 wt% potentially have a strong effect of the apparent viscosity of the mixture, especially with moderate to low density particles.  The actual concentration effect can be influenced by volume percent and particle size.  Small particles are more likely to affect viscosity than larger particles.  Your problem may be more the effect of a high viscosity mixture, including a yield stress, which may mean that the fluid does not move well, or even in all parts of the tank.  If the fluid does not move, you will get stagnant places and settled solids.
 
The solution to both improved solids suspension (if that is the problem) or better handling of a high viscosity (or non-Newtonian, yield stress) fluid is two impellers and larger impellers.  If the current impeller is a hydrofoil with three narrow blades, operating at 340 rpm, it is using only about 0.75 hp of the 7.5 hp motor.  This mixer could work with two (2) 450 mm diameter impellers, drawing about 1.8 hp, or two (2) 500 mm impellers, drawing about 5.2 hp.  Two larger impellers would make better use of the available mixer power and increase the mixing for both solids suspension and high viscosity.  Even if the estimates for power requirements are high, lowering the operating speed of the mixer will solve that problem.
 
This answer contains a number of assumptions about what is happening, but options for changes should make improvements.

The answers by this expert are based on the best available interpretation of the information provided. The consequences of the application of this information are the responsibility of the user. If clarification is needed, please submit a further question.